index.json

[{"authors":["HuaXu"],"categories":null,"content":"","date":1658219044,"expirydate":-62135596800,"kind":"term","lang":"en","lastmod":1658219044,"objectID":"ea83c3d4ee291b52424d8790b81f3023","permalink":"https://thu-xuhua.github.io/author/hua-xu/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/author/hua-xu/","section":"authors","summary":"","tags":null,"title":"Hua Xu","type":"authors"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1647762209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1647762209,"objectID":"33c49a526be2f7911d133234aed7b8a1","permalink":"https://thu-xuhua.github.io/monograph/intent/","publishdate":"2022-03-20T15:43:29+08:00","relpermalink":"/monograph/intent/","section":"monograph","summary":"Inclusive robots can interact naturally with the work environment, humans, and other robots, adapt to complex dynamic environments autonomously, and work collaboratively. [Keen and considerate] natural interaction is one of the hot research issues of inclusive service robots. At present, there is an urgent need for robots and humans to have the ability to understand the intention of interactive dialogue. This book is based on the field of human-computer understanding based on deep learning methods. Starting from the knowledge of human-computer dialogue intentions, it systematically introduces intention recognition, unknown intention detection, and new intention discovery in human-computer dialogue. This book is the first domestic professional book to present interactive dialogue intention analysis in inclusive robot natural interaction. It can help readers master the key technologies and basic knowledge of human-machine dialogue intention analysis in inclusive robot research and track the development frontiers in this field. Provide meaningful learning and research references.","tags":[],"title":"Natural Interaction for Tri-Co Robots (1) Human-machine Dialogue Intention Understanding","type":"monograph"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1646207009,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1646207009,"objectID":"ac3ff13e61fb460d5b7457abbbb13f72","permalink":"https://thu-xuhua.github.io/textbook/datamining-methodandapplication2/","publishdate":"2022-03-02T15:43:29+08:00","relpermalink":"/textbook/datamining-methodandapplication2/","section":"textbook","summary":"Mainly based on the teaching practice and accumulation of the Data Mining Methods and Applications course set up by Tsinghua University, referring to the teaching system of relevant courses of famous foreign universities in recent years, systematically introducing the basic concepts and basic principles of data mining; combining some typical applications Examples show general patterns and ideas for solving problems with data mining thinking methods.","tags":["Data Mining"],"title":"Data Mining: Methodology and Applications(2nd edition)","type":"textbook"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1409557409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1409557409,"objectID":"aa585aa49362c42e9f2830d01bcf8f37","permalink":"https://thu-xuhua.github.io/textbook/datamining-methodandapplication/","publishdate":"2014-09-01T15:43:29+08:00","relpermalink":"/textbook/datamining-methodandapplication/","section":"textbook","summary":"Mainly based on the teaching practice and accumulation of the Data Mining Methods and Applications course set up by Tsinghua University, referring to the teaching system of relevant courses of famous foreign universities in recent years, systematically introducing the basic concepts and basic principles of data mining; combining some typical applications Examples show general patterns and ideas for solving problems with data mining thinking methods.","tags":["Data Mining"],"title":"Data Mining: Methodology and Applications","type":"textbook"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1409557409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1409557409,"objectID":"1d2e51d6cd32618b7ac91b67270b41af","permalink":"https://thu-xuhua.github.io/monograph/evolutionary-machine-learning/","publishdate":"2014-09-01T15:43:29+08:00","relpermalink":"/monograph/evolutionary-machine-learning/","section":"monograph","summary":"Mainly discusses the content of this book: it is a learning classifier and feature selection method, the key is to do both of the integration of research, will study the classification of the classifier model building process and feature selection of feature subset unified integration based on the genetic search process of machine learning framework, at the same time improve the prediction performance of classification algorithm and operation efficiency; Secondly, a learning classifier based on distribution estimation algorithm is introduced to improve the searching quality of rule space. This book can be used as a teaching material and reference book for big data and artificial intelligence majors.","tags":["Data Mining"],"title":"Evolutionary Machine Learning","type":"monograph"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1201851809,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1201851809,"objectID":"2bfeedb392926ea9bc043feb5ad0e8ed","permalink":"https://thu-xuhua.github.io/monograph/petrinettheoryandapplications/","publishdate":"2008-02-01T15:43:29+08:00","relpermalink":"/monograph/petrinettheoryandapplications/","section":"monograph","summary":"Chapter 12: Timed Hierarchical Object-oriented Petri Net, I-Tech Education and Publishing, Vienna, Austria, 2008, pp.253-280, ISNN:978-3-902613-12-7 (Hua Hu participated in the writing, published in February 2008)","tags":[],"title":"Petri Net: Theory and Applications","type":"monograph"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1504251809,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1504251809,"objectID":"cb3a9f5f79958edae4960ecccacd7726","permalink":"https://thu-xuhua.github.io/textbook/datamining-methodandapplication-case/","publishdate":"2017-09-01T15:43:29+08:00","relpermalink":"/textbook/datamining-methodandapplication-case/","section":"textbook","summary":"Mainly based on the teaching practice and accumulation of the Data Mining Methods and Applications course set up by Tsinghua University, referring to the teaching system of relevant courses of famous foreign universities in recent years, systematically introducing the basic concepts and basic principles of data mining; combining some typical applications Examples show general patterns and ideas for solving problems with data mining thinking methods.","tags":["Data Mining"],"title":"Data Mining: Methods and Applications - Application Cases","type":"textbook"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1470037409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1470037409,"objectID":"f3c91f32b83aaa1a5de73ea397ed32b4","permalink":"https://thu-xuhua.github.io/textbook/bigdatatechnologyandindustryapplications/","publishdate":"2016-08-01T15:43:29+08:00","relpermalink":"/textbook/bigdatatechnologyandindustryapplications/","section":"textbook","summary":"How to define big data? How to apply big data? What is big data thinking? How to learn big data? How to build a big data platform? How to apply big data in the industry? This series of problems are very confusing problems in the current era of big data boom. Big Data Technology and Industry Application faces these questions directly, answers the above questions from the perspective of practitioners, and hopes to provide some help to beginners in the big data industry.","tags":["Big Data"],"title":"Big Data Technology and Industry Applications","type":"textbook"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1454312609,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1454312609,"objectID":"4949ddee0a08346cb8809beaaa742b23","permalink":"https://thu-xuhua.github.io/monograph/sentimentanalysisandontologyengineering/","publishdate":"2016-02-01T15:43:29+08:00","relpermalink":"/monograph/sentimentanalysisandontologyengineering/","section":"monograph","summary":"Chapter 10: Chinese Micro-Blog Emotion Classification by Exploiting Linguistic Features and SVMperf ), Springer International Publishing, 2016, pp. 221-236, ISNN:978-3-319-30317-8 (Hua Hu participated in the writing, published in February 2016)","tags":[],"title":"Sentiment Analysis and Ontology Engineering","type":"monograph"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1201851809,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1201851809,"objectID":"2c5c9a0907335dbc37fce7e6ceebe45b","permalink":"https://thu-xuhua.github.io/monograph/recentadvancesinmulti-robotsystems/","publishdate":"2008-02-01T15:43:29+08:00","relpermalink":"/monograph/recentadvancesinmulti-robotsystems/","section":"monograph","summary":"Chapter 13: A Novel Modeling Method for Cooperative Multi-robot Systems Using Fuzzy Timed Agent Based Petri Nets ), I-Tech Education and Publishing, Vienna, Austria, 2008, pp.249-262, ISNN:978-3-902613-24-0 (Hua Hu participated in the writing, published in February 2008)","tags":[],"title":"Recent Advances in Multi-robot Systems","type":"monograph"},{"authors":["JunhuiDeng"],"categories":[],"content":"","date":1107243809,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1107243809,"objectID":"30837e0151d041aff588987056d3a925","permalink":"https://thu-xuhua.github.io/textbook/%E8%AE%A1%E7%AE%97%E5%87%A0%E4%BD%95%E7%AE%97%E6%B3%95%E4%B8%8E%E5%BA%94%E7%94%A8/","publishdate":"2005-02-01T15:43:29+08:00","relpermalink":"/textbook/%E8%AE%A1%E7%AE%97%E5%87%A0%E4%BD%95%E7%AE%97%E6%B3%95%E4%B8%8E%E5%BA%94%E7%94%A8/","section":"textbook","summary":"The first four chapters of \"Computational Geometry: Algorithms and Applications (3rd Edition)\" discuss geometric algorithms, including geometric intersection, triangulation, linear programming, etc. The random algorithm involved is also \"Computational Geometry: Algorithms and Applications (Third Edition)\" a distinctive feature. Chapters 5 to 10 introduce a variety of geometric structures. Chapters 11 to 16 continue to discuss several geometric algorithms and their data structures based on practical problems, they are also further deepening of the content of the first 10 chapters. \"Computational Geometry: Algorithms and Applications (3rd Edition)\" is not only comprehensive in content, but also closely related to practical applications, with prominent points. It has in-depth explanations, and each chapter has \"notes and comments\" and \"exercises\" for the convenience of readers A deeper understanding has been used as teaching materials by many universities around the world.","tags":[],"title":"Computational Geometry: Algorithms and Applications","type":"textbook"},{"authors":["JunhuiDeng"],"categories":[],"content":"","date":1138779809,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1138779809,"objectID":"c70ac8a450da7476c774fb168d54bc54","permalink":"https://thu-xuhua.github.io/textbook/%E6%95%B0%E6%8D%AE%E7%BB%93%E6%9E%84%E4%B8%8E%E7%AE%97%E6%B3%95java%E8%AF%AD%E8%A8%80%E6%8F%8F%E8%BF%B0/","publishdate":"2006-02-01T15:43:29+08:00","relpermalink":"/textbook/%E6%95%B0%E6%8D%AE%E7%BB%93%E6%9E%84%E4%B8%8E%E7%AE%97%E6%B3%95java%E8%AF%AD%E8%A8%80%E6%8F%8F%E8%BF%B0/","section":"textbook","summary":"This book fully demonstrates the application of object-oriented technology in modern data structure theory, generally using abstraction, encapsulation, and inheritance technologies. This book not only introduces the basic data structure, but also introduces the application, implementation and analysis methods of the algorithm in combination with specific problems. The book also unifies various graph algorithms through the traversal algorithm framework and implements it based on the traversal algorithm template. Unique among similar textbooks.","tags":[],"title":"Data Structures and Algorithms (Java Description)","type":"textbook"},{"authors":["JunhuiDeng"],"categories":[],"content":"","date":1075621409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1075621409,"objectID":"7dedbd79296664918c876247a7356cba","permalink":"https://thu-xuhua.github.io/monograph/%E8%B4%A8%E9%87%8F%E8%BD%AF%E4%BB%B6%E7%AE%A1%E7%90%86/","publishdate":"2004-02-01T15:43:29+08:00","relpermalink":"/monograph/%E8%B4%A8%E9%87%8F%E8%BD%AF%E4%BB%B6%E7%AE%A1%E7%90%86/","section":"monograph","summary":"Tsinghua University Press (Jun. 2004) ISBN: 7-302-08298-7 （Original Work：Gerald M. Weinberg, Quality Software Management:Systems Thinking ,Dorset House (Sep. 1991), ISBN: 0-932-63322-6.）","tags":[],"title":"Quality Software Management (Volume 1)-System Thinking","type":"monograph"},{"authors":["JunhuiDeng"],"categories":[],"content":"","date":1044085409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1044085409,"objectID":"edca37e22f48ff86ab5a6e634b5b3a8a","permalink":"https://thu-xuhua.github.io/monograph/%E7%A8%8B%E5%BA%8F%E5%BC%80%E5%8F%91%E5%BF%83%E7%90%86%E5%AD%A6%E9%93%B6%E5%B9%B4%E7%BA%AA%E5%BF%B5%E7%89%88/","publishdate":"2003-02-01T15:43:29+08:00","relpermalink":"/monograph/%E7%A8%8B%E5%BA%8F%E5%BC%80%E5%8F%91%E5%BF%83%E7%90%86%E5%AD%A6%E9%93%B6%E5%B9%B4%E7%BA%AA%E5%BF%B5%E7%89%88/","section":"monograph","summary":"Tsinghua University Press (Sep. 2003) ISBN: 7-302-07026-1 （Original Work：Gerald M. Weinberg, The Psychology of Computer Programming: Silver Anniversary Edition, Dorset House (Sep. 1998), ISBN: 0-932-63342-0.）","tags":[],"title":"The Psychology of Computer Programming","type":"monograph"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1602301409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1602301409,"objectID":"d9809ef0bb72c0fdeb8f093635124ba3","permalink":"https://thu-xuhua.github.io/textbook/%E6%96%87%E6%9C%AC%E5%A4%A7%E6%95%B0%E6%8D%AE%E6%83%85%E6%84%9F%E5%88%86%E6%9E%90/","publishdate":"2020-10-10T11:43:29+08:00","relpermalink":"/textbook/%E6%96%87%E6%9C%AC%E5%A4%A7%E6%95%B0%E6%8D%AE%E6%83%85%E6%84%9F%E5%88%86%E6%9E%90/","section":"textbook","summary":"This book explains the text big data sentiment analysis technology from multiple dimensions. The content covers natural language processing and text emotion and sentiment methods, as well as sentiment analysis of Weibo texts and their incentives, position analysis of topic-oriented comments, texts Representation method and its application in emotion classification. The book is organized in a modular manner, with strong theoretical and clear organization. The author team implemented the main methods in the book with a serious and rigorous scientific attitude, and described the effects of various methods. This book can provide help for the study and scientific research work of college students in related majors (such as computer science and technology, software engineering, etc.). It also has a higher reference value for engineering and technical personnel engaged in text mining and natural language processing.","tags":[],"title":"Big Data Sentiment Analysis in Text","type":"textbook"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1582271009,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1582271009,"objectID":"a400bd5cb2ce6d457b1dc926cbb90884","permalink":"https://thu-xuhua.github.io/patent/3000-based-on-the-product-review-sensibility-classification-method-and-device-that-improve-svms/","publishdate":"2020-02-21T15:43:29+08:00","relpermalink":"/patent/3000-based-on-the-product-review-sensibility-classification-method-and-device-that-improve-svms/","section":"patent","summary":"","tags":[],"title":"Based on the product review sensibility classification method and device that improve SVMs","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1571643809,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1571643809,"objectID":"584f8922a22246e7256b51d9879f35ae","permalink":"https://thu-xuhua.github.io/patent/3001-banks-background-task-runs-the-correlation-analysis-and-device-of-batch-time/","publishdate":"2019-10-21T15:43:29+08:00","relpermalink":"/patent/3001-banks-background-task-runs-the-correlation-analysis-and-device-of-batch-time/","section":"patent","summary":"","tags":[],"title":"Bank's background task runs the correlation analysis and device of batch time","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1565336609,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1565336609,"objectID":"39953ee7c3f6a97e6e7a6f8272f1d02f","permalink":"https://thu-xuhua.github.io/patent/3002-the-many-emotions-sorting-technique-extracted-based-on-weight-study-and-multiple-features-and-device/","publishdate":"2019-08-09T15:43:29+08:00","relpermalink":"/patent/3002-the-many-emotions-sorting-technique-extracted-based-on-weight-study-and-multiple-features-and-device/","section":"patent","summary":"","tags":[],"title":"The many emotions sorting technique extracted based on weight study and multiple features and device","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1562053409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1562053409,"objectID":"b28d3a58f4d0731829c25bb90d85e70d","permalink":"https://thu-xuhua.github.io/patent/3003-theme-model-based-self-extendable-type-feature-selecting-method-and-system/","publishdate":"2019-07-02T15:43:29+08:00","relpermalink":"/patent/3003-theme-model-based-self-extendable-type-feature-selecting-method-and-system/","section":"patent","summary":"","tags":[],"title":"Theme model based self-extendable type feature selecting method and system","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1559288609,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1559288609,"objectID":"2302e005406548a71ff7528d85982c62","permalink":"https://thu-xuhua.github.io/patent/3004-the-sampling-accelerated-method-of-biterm-topic-models/","publishdate":"2019-05-31T15:43:29+08:00","relpermalink":"/patent/3004-the-sampling-accelerated-method-of-biterm-topic-models/","section":"patent","summary":"","tags":[],"title":"The sampling accelerated method of Biterm topic models","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1549352609,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1549352609,"objectID":"161b668d766fd93e79fc6498b83a4243","permalink":"https://thu-xuhua.github.io/patent/3005-microblog-social-network-based-topic-automated-recommendation-method-and-system/","publishdate":"2019-02-05T15:43:29+08:00","relpermalink":"/patent/3005-microblog-social-network-based-topic-automated-recommendation-method-and-system/","section":"patent","summary":"","tags":[],"title":"Microblog social network based topic automated recommendation method and system","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1549352609,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1549352609,"objectID":"8ad2c572a0165ea290d004edcf962cba","permalink":"https://thu-xuhua.github.io/patent/3006-text-summarization-generation-system-and-method-based-on-coding-decoding-deep-neural-networks/","publishdate":"2019-02-05T15:43:29+08:00","relpermalink":"/patent/3006-text-summarization-generation-system-and-method-based-on-coding-decoding-deep-neural-networks/","section":"patent","summary":"","tags":[],"title":"Text summarization generation system and method based on coding-decoding deep neural networks","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1538466209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1538466209,"objectID":"c5c8c4ba3b6ab1af506318266afdc4c3","permalink":"https://thu-xuhua.github.io/patent/3007-chinese-microblogging-mood-sorting-technique-based-on-markov-logical-network-and-system/","publishdate":"2018-10-02T15:43:29+08:00","relpermalink":"/patent/3007-chinese-microblogging-mood-sorting-technique-based-on-markov-logical-network-and-system/","section":"patent","summary":"","tags":[],"title":"Chinese microblogging mood sorting technique based on Markov logical network and system","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1523346209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1523346209,"objectID":"571f6fb880d2bbc3c3cfee5dd0844ddd","permalink":"https://thu-xuhua.github.io/patent/3008-context-sentiment-classification-method-and-system-based-on-user-comment-text/","publishdate":"2018-04-10T15:43:29+08:00","relpermalink":"/patent/3008-context-sentiment-classification-method-and-system-based-on-user-comment-text/","section":"patent","summary":"","tags":[],"title":"Context sentiment classification method and system based on user comment text","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1518162209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1518162209,"objectID":"7210f049926f6990261d1407ebb416e1","permalink":"https://thu-xuhua.github.io/patent/3009-tps-trading-volumes-predict-the-method-for-visualizing-and-system-with-threshold-range-prediction/","publishdate":"2018-02-09T15:43:29+08:00","relpermalink":"/patent/3009-tps-trading-volumes-predict-the-method-for-visualizing-and-system-with-threshold-range-prediction/","section":"patent","summary":"","tags":[],"title":"TPS trading volumes predict the method for visualizing and system with threshold range prediction","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1576741409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1576741409,"objectID":"46b22e8046d6a066d05f035bba167e31","permalink":"https://thu-xuhua.github.io/patent/3010-method-and-system-for-excavating-interest-of-microblog-users-based-on-lda-theme-model/","publishdate":"2019-12-19T15:43:29+08:00","relpermalink":"/patent/3010-method-and-system-for-excavating-interest-of-microblog-users-based-on-lda-theme-model/","section":"patent","summary":"","tags":[],"title":"Method and system for excavating interest of microblog users based on LDA theme model","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1511250209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1511250209,"objectID":"1ac9f75f7fdd2252620654bef0b31f09","permalink":"https://thu-xuhua.github.io/patent/3011-microblog-user-interest-finding-method-and-system-based-on-microblog-texts/","publishdate":"2017-11-21T15:43:29+08:00","relpermalink":"/patent/3011-microblog-user-interest-finding-method-and-system-based-on-microblog-texts/","section":"patent","summary":"","tags":[],"title":"Microblog user interest finding method and system based on microblog texts","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1508831009,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1508831009,"objectID":"a70f3aad571509ba8cf8a5646fc764b4","permalink":"https://thu-xuhua.github.io/patent/3012-banks-backstage-tps-transaction-events-trend-forecasting-method-and-system/","publishdate":"2017-10-24T15:43:29+08:00","relpermalink":"/patent/3012-banks-backstage-tps-transaction-events-trend-forecasting-method-and-system/","section":"patent","summary":"","tags":[],"title":"Bank's backstage TPS transaction events trend forecasting method and system","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1506066209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1506066209,"objectID":"9ff4b0c6673f2448cfed7a6302624882","permalink":"https://thu-xuhua.github.io/patent/3013-emotion-classification-and-emotion-component-analyzing-method-and-system-based-on-microblog-texts/","publishdate":"2017-09-22T15:43:29+08:00","relpermalink":"/patent/3013-emotion-classification-and-emotion-component-analyzing-method-and-system-based-on-microblog-texts/","section":"patent","summary":"","tags":[],"title":"Emotion classification and emotion component analyzing method and system based on microblog texts","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1492587809,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1492587809,"objectID":"9420d4f1790c4163ef6db35ee565edab","permalink":"https://thu-xuhua.github.io/patent/3014-microblog-text-level-subject-finding-method-and-system-based-on-seed-words/","publishdate":"2017-04-19T15:43:29+08:00","relpermalink":"/patent/3014-microblog-text-level-subject-finding-method-and-system-based-on-seed-words/","section":"patent","summary":"","tags":[],"title":"Microblog text level subject finding method and system based on seed words","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1458719009,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1458719009,"objectID":"faca22998cdbf71cbe2c046b4e6b886f","permalink":"https://thu-xuhua.github.io/patent/3015-the-dispatching-method-of-multiple-robot-devices-of-film-trasport-cooperation-and-system/","publishdate":"2016-03-23T15:43:29+08:00","relpermalink":"/patent/3015-the-dispatching-method-of-multiple-robot-devices-of-film-trasport-cooperation-and-system/","section":"patent","summary":"","tags":[],"title":"The dispatching method of multiple robot devices of film trasport cooperation and system","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1449042209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1449042209,"objectID":"22aab9524c3af7764e8008ec62d32bd0","permalink":"https://thu-xuhua.github.io/patent/3016-parameter-variation-law-simulation-system-oriented-to-function-simulation-of-semiconductor-manufacturing-apparatus/","publishdate":"2015-12-02T15:43:29+08:00","relpermalink":"/patent/3016-parameter-variation-law-simulation-system-oriented-to-function-simulation-of-semiconductor-manufacturing-apparatus/","section":"patent","summary":"","tags":[],"title":"Parameter variation law simulation system oriented to function simulation of semiconductor manufacturing apparatus","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1439365409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1439365409,"objectID":"4733a52a2fb2daad7625f47874f811ce","permalink":"https://thu-xuhua.github.io/patent/3017-a-kind-of-skin-forecasting-methodology-and-prognoses-system-thereof/","publishdate":"2015-08-12T15:43:29+08:00","relpermalink":"/patent/3017-a-kind-of-skin-forecasting-methodology-and-prognoses-system-thereof/","section":"patent","summary":"","tags":[],"title":"A kind of skin Forecasting Methodology and prognoses system thereof","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1397029409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1397029409,"objectID":"6fe82ee8f3eaca1e7236bf77523e7924","permalink":"https://thu-xuhua.github.io/patent/3018-general-control-kernel-system-used-for-integrated-circuit-manufacturing-equipment/","publishdate":"2014-04-09T15:43:29+08:00","relpermalink":"/patent/3018-general-control-kernel-system-used-for-integrated-circuit-manufacturing-equipment/","section":"patent","summary":"","tags":[],"title":"General control kernel system used for integrated circuit manufacturing equipment","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1389167009,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1389167009,"objectID":"a7acb0f89fb11dc01205226d17753e60","permalink":"https://thu-xuhua.github.io/patent/3019-method-and-device-for-configuring-and-parsing-dynamic-protocol/","publishdate":"2014-01-08T15:43:29+08:00","relpermalink":"/patent/3019-method-and-device-for-configuring-and-parsing-dynamic-protocol/","section":"patent","summary":"","tags":[],"title":"Method and device for configuring and parsing dynamic protocol","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1375256609,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1375256609,"objectID":"aa432e7909c2e460d5df0718e3ccc3f6","permalink":"https://thu-xuhua.github.io/patent/3020-universal-control-system-for-integrated-circuit-manufacturing-equipment/","publishdate":"2013-07-31T15:43:29+08:00","relpermalink":"/patent/3020-universal-control-system-for-integrated-circuit-manufacturing-equipment/","section":"patent","summary":"","tags":[],"title":"Universal control system for integrated circuit manufacturing equipment","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1371023009,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1371023009,"objectID":"99e3b079e9bcfa6a08c41ffe113f0f82","permalink":"https://thu-xuhua.github.io/patent/3022-semiconductor-manufacturing-equipment-functional-simulation-oriented-communication-protocol-mode-configuration-method/","publishdate":"2013-06-12T15:43:29+08:00","relpermalink":"/patent/3022-semiconductor-manufacturing-equipment-functional-simulation-oriented-communication-protocol-mode-configuration-method/","section":"patent","summary":"","tags":[],"title":"Semiconductor manufacturing equipment functional simulation-oriented communication protocol mode configuration method","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1369208609,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1369208609,"objectID":"d063efd7f7288d4d4fefb1105b814451","permalink":"https://thu-xuhua.github.io/patent/3021-internal-interlocking-management-method-and-device-for-simulation-system/","publishdate":"2013-05-22T15:43:29+08:00","relpermalink":"/patent/3021-internal-interlocking-management-method-and-device-for-simulation-system/","section":"patent","summary":"","tags":[],"title":"Internal interlocking management method and device for simulation system","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1365579809,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1365579809,"objectID":"107f02410f8f14416f78861d2555b778","permalink":"https://thu-xuhua.github.io/patent/3023-parallelized-simulation-multithread-management-method/","publishdate":"2013-04-10T15:43:29+08:00","relpermalink":"/patent/3023-parallelized-simulation-multithread-management-method/","section":"patent","summary":"","tags":[],"title":"Parallelized simulation multithread management method","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1364370209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1364370209,"objectID":"6d199d0e378ed4f6b191d8f653ae108e","permalink":"https://thu-xuhua.github.io/patent/3024-method-and-apparatus-for-predicting-antibiotic-performance-or-critical-ingredient-content-of-detergent/","publishdate":"2013-03-27T15:43:29+08:00","relpermalink":"/patent/3024-method-and-apparatus-for-predicting-antibiotic-performance-or-critical-ingredient-content-of-detergent/","section":"patent","summary":"","tags":[],"title":"Method and apparatus for predicting antibiotic performance or critical ingredient content of detergent","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1355903009,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1355903009,"objectID":"6fcc2d8db93eefe16c9c363a9c215001","permalink":"https://thu-xuhua.github.io/patent/3025-communication-mode-management-method-and-system-for-emulating-semiconductor-manufacturing-equipment/","publishdate":"2012-12-19T15:43:29+08:00","relpermalink":"/patent/3025-communication-mode-management-method-and-system-for-emulating-semiconductor-manufacturing-equipment/","section":"patent","summary":"","tags":[],"title":"Communication mode management method and system for emulating semiconductor manufacturing equipment","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1349855009,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1349855009,"objectID":"86539553c05c80ab42166810b4a66b05","permalink":"https://thu-xuhua.github.io/patent/3026-method-and-device-for-constructing-and-managing-prototype-device-library/","publishdate":"2012-10-10T15:43:29+08:00","relpermalink":"/patent/3026-method-and-device-for-constructing-and-managing-prototype-device-library/","section":"patent","summary":"","tags":[],"title":"Method and device for constructing and managing prototype device library","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1342597409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1342597409,"objectID":"2ecf3d0c27da4dbc07ed8c778edbbe3c","permalink":"https://thu-xuhua.github.io/patent/3027-simulated-batch-configuration-document-management-method-oriented-to-semiconductor-manufacturing-equipment-function/","publishdate":"2012-07-18T15:43:29+08:00","relpermalink":"/patent/3027-simulated-batch-configuration-document-management-method-oriented-to-semiconductor-manufacturing-equipment-function/","section":"patent","summary":"","tags":[],"title":"Simulated batch configuration document management method oriented to semiconductor manufacturing equipment function","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1227685409,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1227685409,"objectID":"b6e91daa37dc3fb7ea1df372d984d255","permalink":"https://thu-xuhua.github.io/patent/3028-method-and-device-for-storing-and-managing-configuration-data-of-simulation-system/","publishdate":"2008-11-26T15:43:29+08:00","relpermalink":"/patent/3028-method-and-device-for-storing-and-managing-configuration-data-of-simulation-system/","section":"patent","summary":"","tags":[],"title":"Method and device for storing and managing configuration data of simulation system","type":"patent"},{"authors":["Hua Xu"],"categories":[],"content":"","date":1225266209,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1225266209,"objectID":"47bf33fe004484eea64b665d0c4d1c0c","permalink":"https://thu-xuhua.github.io/patent/3029-etchingmachine-cluster-controller-and-process-module-controller-communication-system-and-method/","publishdate":"2008-10-29T15:43:29+08:00","relpermalink":"/patent/3029-etchingmachine-cluster-controller-and-process-module-controller-communication-system-and-method/","section":"patent","summary":"","tags":[],"title":"Etchingmachine cluster controller and process module controller communication system and method","type":"patent"},{"authors":["ZhijingWu","Hua Xu","JingliangFang","KaiGao"],"categories":[],"content":"","date":1658219044,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1658219044,"objectID":"ead9f2870e66b19f8d91ce94524fa012","permalink":"https://thu-xuhua.github.io/publication/continual-machine-reading-comprehension-via-uncertainty-aware-fixed-memory-and-adversarial-domain-adaptation/","publishdate":"2022-07-19T16:24:04+08:00","relpermalink":"/publication/continual-machine-reading-comprehension-via-uncertainty-aware-fixed-memory-and-adversarial-domain-adaptation/","section":"publication","summary":"Continual Machine Reading Comprehension aims to incrementally learn from a continuous data stream across time without access the previous seen data, which is crucial for the development of real-world MRC systems. However, it is a great challenge to learn a new domain incrementally without catastrophically forgetting previous knowledge. In this paper, MA-MRC, a continual MRC model with uncertainty-aware fixed Memory and Adversarial domain adaptation, is proposed. In MA-MRC, a fixed size memory stores a small number of samples in previous domain data along with an uncertainty-aware updating strategy when new domain data arrives. For incremental learning, MA-MRC not only keeps a stable understanding by learning both memory and new domain data, but also makes full use of the domain adaptation relationship between them by adversarial learning strategy. The experimental results show that MA-MRC is superior to strong baselines and has a substantial incremental learning ability without catastrophically forgetting under two different continual MRC settings.","tags":[],"title":"Continual Machine Reading Comprehension via Uncertainty-aware Fixed Memory and Adversarial Domain Adaptation ","type":"publication"},{"authors":["LiChen","Hua Xu"],"categories":[],"content":"","date":1658219044,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1658219044,"objectID":"e609e769d5d1a86ae3e279650919b148","permalink":"https://thu-xuhua.github.io/publication/mfenas/","publishdate":"2022-07-19T16:24:04+08:00","relpermalink":"/publication/mfenas/","section":"publication","summary":"Neural Architecture Search (NAS) aims to automatically find neural network architectures competitive with human-designed ones. Despite the remarkable progress achieved, existing NAS methods still suffer from vast computational resources cost. Inspired by MFEA, we model the NAS task as a two-factorial problem and propose a multifactorial evolutionary neural architecture search (MFENAS) algorithm to solve it. MFENAS divides a population into two subgroups according to factors, and then the factors influence the evolution and knowledge transfer between subgroups. Experimental results of NATS-Bench demonstrate the efficiency of the proposed MFENAS in finding optimal structures under resource constraints compared to other state-of-the-art methods.","tags":[],"title":"MFENAS: Multifactorial Evolution for Neural Architecture Search","type":"publication"},{"authors":["WangHongyan","XuHua","YuanYuan"],"categories":[],"content":"","date":1651393444,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1651393444,"objectID":"505d3cfdcb4ae94233ab44aa73a9dfc5","permalink":"https://thu-xuhua.github.io/publication/high-dimensional-expensive-multi-objective-optimization-via-additive-structure/","publishdate":"2022-05-01T16:24:04+08:00","relpermalink":"/publication/high-dimensional-expensive-multi-objective-optimization-via-additive-structure/","section":"publication","summary":"Expensive multi-objective problems (MOPs) are extremely challenging due to the high evaluation cost to find satisfying solutions with adequate precision, especially in high-dimensional cases. However, most of the current EGO-based algorithms for expensive MOPs are limited to low decision dimensions because of the exponential difficulty in high dimensional circumstances. This paper presents High-Dimensional Expensive Multi-objective Optimization with Additive structure (ADD-HDEMO) to solve high-dimensional expensive MOPs via additive structural kernel and identifies two key challenges in this endeavor. First, we integrate multiple sub-objectives in high-dimensional expensive MOPs into a single objective with the decision space unchanged. Then, we infer the dependence correlation between the decision and objective space of the augmented EMOP via an additive GP kernel structure where Gibbs sampling is used to learn the latent additive structure. Furthermore, we parallel the proposed algorithm by introducing a multi-point sampling mechanism when recommending infill points. The effectiveness of the proposed method is evaluated on ZDT and DTLZ benchmarks compared with three other EGO-based multi-objective optimization approaches, ParEGO, SMS-EGO and MOEA/D-EGO. Our analyses demonstrate that ADD-HDEMO is effective in solving high-dimensional expensive MOPs.","tags":[],"title":"High-dimensional expensive multi-objective optimization via additive structure","type":"publication"},{"authors":["Huisheng Mao","Ziqi Yuan","Hua Xu","Wenmeng Yu","Yihe Liu","Kai Gao"],"categories":[],"content":"","date":1647415444,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1647415444,"objectID":"71f77290e373f601482fc2c9952ad45c","permalink":"https://thu-xuhua.github.io/publication/m-sena/","publishdate":"2022-03-16T15:24:04+08:00","relpermalink":"/publication/m-sena/","section":"publication","summary":"M-SENA is an open-sourced platform for Multimodal Sentiment Analysis. It aims to facilitate advanced research by providing flexible toolkits, reliable benchmarks, and intuitive demonstrations. The platform features a fully modular video sentiment analysis framework consisting of data management, feature extraction, model training, and result analysis modules. In this paper, we first illustrate the overall architecture of the M-SENA platform and then introduce features of the core modules. Reliable baseline results of different modality features and MSA benchmarks are also reported. Moreover, we use model evaluation and analysis tools provided by M-SENA to present intermediate representation visualization, on-the-fly instance test, and generalization ability test results. ","tags":[],"title":"M-SENA: An Integrated Platform for Multimodal Sentiment Analysis","type":"publication"},{"authors":["Kang Zhao","Hua Xu","Jiangong Yang","Kai Gao"],"categories":[],"content":"","date":1647329044,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1647329044,"objectID":"e04066c112ec77ea309e86090bb7c8e6","permalink":"https://thu-xuhua.github.io/publication/crl/","publishdate":"2022-03-15T15:24:04+08:00","relpermalink":"/publication/crl/","section":"publication","summary":"Continual relation extraction (CRE) aims to continuously train a model on data with new relations while avoiding forgetting old ones. Some previous work has proved that storing a few typical samples of old relations and replaying them when learning new relations can effectively avoid forgetting. However, these memory-based methods tend to overfit the memory samples and perform poorly on imbalanced datasets. To solve these challenges, a consistent representation learning method is proposed, which maintains the stability of the relation embedding by adopting contrastive learning and knowledge distillation when replaying memory. Specifically, supervised contrastive learning based on a memory bank is first used to train each new task so that the model can effectively learn the relation representation. Then, contrastive replay is conducted of the samples in memory and makes the model retain the knowledge of historical relations through memory knowledge distillation to prevent the catastrophic forgetting of the old task. The proposed method can better learn consistent representations to alleviate forgetting effectively. Extensive experiments on FewRel and TACRED datasets show that our method significantly outperforms state-of-theart baselines and yield strong robustness on the imbalanced dataset.","tags":[],"title":"Consistent Representation Learning for Continual Relation Extraction","type":"publication"},{"authors":["Hua Xu","ZiqiYuan","KangZhao","YunfengXu","JiyunZou","KaiGao"],"categories":[],"content":"","date":1644913444,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1644913444,"objectID":"ba0f50233cdcfbb4b47d4b5b4a3bf639","permalink":"https://thu-xuhua.github.io/publication/gar-net/","publishdate":"2022-02-15T16:24:04+08:00","relpermalink":"/publication/gar-net/","section":"publication","summary":"Conversation understanding, as a necessary step for many applications, including social media, education, and argumentation mining, has been gaining increasing attention from the research community. Reasoning over long-term dependent contextual information is the key to utterance-level conversation understanding. Aiming to emphasize the importance of contextual reasoning, an end-to-end graph attention reasoning network which takes both word-level and utterance-level context into concern is proposed. To be specific, a word-level encoder with a novel convolution gate is first built to filter out irrelevant contextual information. Based on the representation extracted by word-level encoder, a graph reasoning network is designed to utilize the context among utterance-level, where the entire conversation is treated as a fully connected graph, utterances as nodes, and attention scores between utterances as edges. The proposed model is a general framework for conversation understanding tasks, which can be flexibly applied on most conversation datasets without changing the network architecture. Furthermore, we revise the tensor fusion network by adding a residual connection to explore cross-modal conversation understanding. For uni-modal scene (text modality), experiments show that the proposed method surpasses current state-of-the-art methods on emotion recognition, intent classification, and dialogue act identification tasks. For cross-modal scenes (text modality and audio modality), experiments on IEMOCAP and MELD datasets show that the proposed method can use cross-modal information to improve model performance.","tags":[],"title":"GAR-Net: A Graph Attention Reasoning Network for Conversation Understanding","type":"publication"},{"authors":["Huisheng Mao","Baozheng Zhang","Hua Xu","Kai Gao"],"categories":[],"content":"","date":1641972244,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1641972244,"objectID":"750b1ac20c18e626cd734ed9133c4929","permalink":"https://thu-xuhua.github.io/publication/an-end-to-end-traditional-chinese-medicine-constitution-assessment-system-based-on-multimodal-clinical-feature-representation-and-fusion/","publishdate":"2022-01-31T15:24:04+08:00","relpermalink":"/publication/an-end-to-end-traditional-chinese-medicine-constitution-assessment-system-based-on-multimodal-clinical-feature-representation-and-fusion/","section":"publication","summary":"Traditional Chinese Medicine (TCM) constitution is a fundamental concept in TCM theory. It is determined by multimodal TCM clinical features which, in turn, are obtained from TCM clinical information of image (face, tongue, etc.), audio (pulse and voice), and text (inquiry) modality. The auto assessment of TCM constitution is faced with two major challenges: (1) learning discriminative TCM clinical feature representations; (2) jointly processing the features using multimodal fusion techniques. The TCM Constitution Assessment System (TCM-CAS) is proposed to provide an end-to-end solution to this task, along with auxiliary functions to aid TCM researchers. To improve the results of TCM constitution prediction, the system combines multiple machine learning algorithms such as facial landmark detection, image segmentation, graph neural networks and multimodal fusion. Extensive experiments are conducted on a four-category multimodal TCM constitution dataset, and the proposed method achieves state-of-the-art accuracy. Provided with datasets containing annotations of diseases, the system can also perform automatic disease diagnosis from a TCM perspective.","tags":[],"title":"An End-to-End Traditional Chinese Medicine Constitution Assessment System Based on Multimodal Clinical Feature Representation and Fusion","type":"publication"},{"authors":["WenmengYu","Hua Xu"],"categories":[],"content":"","date":1635755044,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1635755044,"objectID":"aa2e330cb394dfe7750182253390ced2","permalink":"https://thu-xuhua.github.io/publication/co-attentive-multi-task-convolutional-neural-network-for-facial-expression-recognition/","publishdate":"2021-11-01T16:24:04+08:00","relpermalink":"/publication/co-attentive-multi-task-convolutional-neural-network-for-facial-expression-recognition/","section":"publication","summary":"Previous research on Facial Expression Recognition (FER) assisted by facial landmarks mainly focused on single-task learning or hard-parameter sharing based multi-task learning. However, soft-parameter shar- ing based methods have not been explored in this area. Therefore, this paper adopts Facial Landmark Detection (FLD) as the auxiliary task and explores new multi-task learning strategies for FER. First, three classical multi-task structures, including Hard-Parameter Sharing (HPS), Cross-Stitch Network (CSN), and Partially Shared Multi-task Convolutional Neural Network (PS-MCNN), are used to verify the advantages of multi-task learning for FER. Then, we propose a new end-to-end Co-attentive Multi-task Convolutional Neural Network (CMCNN), which is composed of the Channel Co-Attention Module (CCAM) and the Spa- tial Co-Attention Module (SCAM). Functionally, the CCAM generates the channel co-attention scores by capturing the inter-dependencies of different channels between FER and FLD tasks. The SCAM combines the max- and average-pooling operations to formulate the spatial co-attention scores. Finally, we con- duct extensive experiments on four widely used benchmark facial expression databases, including RAF, SFEW2, CK+, and Oulu-CASIA. Extensive experimental results show that our approach achieves better performance than single-task and multi-task baselines, fully validating multi-task learning s effectiveness and generalizability.","tags":[],"title":"Co-attentive multi-task convolutional neural network for facial expression recognition","type":"publication"},{"authors":["ZiqiYuan","WeiLi","Hua Xu","WenmengYu"],"categories":[],"content":"","date":1634718244,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1634718244,"objectID":"dcb19b7e0d77ae162f602fb808c04194","permalink":"https://thu-xuhua.github.io/publication/transformer-based-feature-reconstruction-network-for-robust-multimodal-sentiment-analysis/","publishdate":"2021-10-20T16:24:04+08:00","relpermalink":"/publication/transformer-based-feature-reconstruction-network-for-robust-multimodal-sentiment-analysis/","section":"publication","summary":"Improving robustness against data missing has become one of the core challenges in Multimodal Sentiment Analysis (MSA), which aims to judge speaker sentiments from the language, visual, and acoustic signals. In the current research, translation-based methods and tensor regularization methods are proposed for MSA with incomplete modality features. However, both of them fail to cope with random modality feature missing in non-aligned sequences. In this paper, a transformer-based feature reconstruction network (TFR-Net) is proposed to improve the robustness of models for the random missing in non-aligned modality sequences. First, intramodal and inter-modal attention-based extractors are adopted to learn robust representations for each element in modality sequences. Then, a reconstruction module is proposed to generate the missing modality features. With the supervision of SmoothL1Loss between generated and complete sequences, TFR-Net is expected to learn semantic-level features corresponding to missing features. Extensive experiments on two public benchmark datasets show that our model achieves good results against data missing across various missing modality combinations and various missing degrees.","tags":[],"title":"Transformer-based Feature Reconstruction Network for Robust Multimodal Sentiment Analysis","type":"publication"},{"authors":["HanleiZhang","XiaotengLi","Hua Xu","PanpanZhang","KangZhao","KaiGao"],"categories":[],"content":"","date":1620462244,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1620462244,"objectID":"e506f9d300b6b691d19ea82504a0d692","permalink":"https://thu-xuhua.github.io/publication/textoir/","publishdate":"2021-05-08T16:24:04+08:00","relpermalink":"/publication/textoir/","section":"publication","summary":"TEXTOIR is the first integrated and visualized platform for text open intent recognition. It is composed of two main modules: open intent detection and open intent discovery. Each module integrates most of the state-of-the-art algorithms and benchmark intent datasets. It also contains an overall framework connecting the two modules in a pipeline scheme. In addition, this platform has visualized tools for data and model management, training, evaluation and analysis of the performance from different aspects. TEXTOIR provides useful toolkits and convenient visualized interfaces for each sub-module , and designs a framework to implement a complete process to both identify known intents and discover open intents. Codes can be found at https://github.com/thuiar/TEXTOIR","tags":[],"title":"TEXTOIR: An Integrated and Visualized Platform for Text Open Intent Recognition","type":"publication"},{"authors":["HanleiZhang","Hua Xu","TingenLin","RuiLv"],"categories":[],"content":"","date":1613558326,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1613558326,"objectID":"ec89d89560bb97f9139ac0639d274dad","permalink":"https://thu-xuhua.github.io/publication/deepaligned-clustering/","publishdate":"2021-02-17T18:38:46+08:00","relpermalink":"/publication/deepaligned-clustering/","section":"publication","summary":"Discovering new intents is a crucial task in dialogue systems. Most existing methods are limited in transferring the prior knowledge from known intents to new intents. These methods also have difficulties in providing high-quality supervised signals to learn clustering-friendly features for grouping unlabeled intents. In this work, we propose an effective method (Deep Aligned Clustering) to discover new intents with the aid of limited known intent data. Firstly, we leverage a few labeled known intent samples as prior knowledge to pre-train the model. Then, we perform k-means to produce cluster assignments as pseudo-labels. Moreover, we propose an alignment strategy to tackle the label inconsistency problem during clustering assignments. Finally, we learn the intent representations under the supervision of the aligned pseudo-labels. With an unknown number of new intents, we predict the number of intent categories by eliminating lowconfidence intent-wise clusters. Extensive experiments on two benchmark datasets show that our method is more robust and achieves substantial improvements over the state-of-theart methods.","tags":[],"title":"Discovering New Intents with Deep Aligned Clustering","type":"publication"},{"authors":["HanleiZhang","Hua Xu","TingenLin"],"categories":[],"content":"","date":1613471926,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1613471926,"objectID":"e37b2d1ba3f5c8ec85d8e0c153e25fd7","permalink":"https://thu-xuhua.github.io/publication/adaptive-decision-boundary/","publishdate":"2021-02-16T18:38:46+08:00","relpermalink":"/publication/adaptive-decision-boundary/","section":"publication","summary":"Open intent classification is a challenging task in dialogue systems. On the one hand, we should ensure the classification quality of known intents. On the other hand, we need to identify the open (unknown) intent during testing. Current models are limited in finding the appropriate decision boundary to balance the performances of both known and open intents. In this paper, we propose a post-processing method to learn the adaptive decision boundary (ADB) for open intent classification. We first utilize the labeled known intent samples to pre-train the model. Then, we use the well-trained features to automatically learn the adaptive spherical decision boundaries for each known intent. Specifically, we propose a new loss function to balance both the empirical risk and the open space risk. Our method does not need open samples and is free from modifying the model architecture. We find our approach is surprisingly insensitive with less labeled  data and fewer known intents. Extensive experiments on three benchmark datasets show that our method yields significant improvements compared with the state-of-the-art methods.","tags":[],"title":"Deep Open Intent Classification with Adaptive Decision Boundary","type":"publication"},{"authors":["WenmengYu","Hua Xu","ZiqiYuan","JieleWu"],"categories":[],"content":"","date":1613385526,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1613385526,"objectID":"54974a840628e4cbcc4a43055bc6d6c2","permalink":"https://thu-xuhua.github.io/publication/self-mm/","publishdate":"2021-02-15T18:38:46+08:00","relpermalink":"/publication/self-mm/","section":"publication","summary":"Representation Learning is a significant and challenging task in multimodal learning. Effective modality representations should contain two parts of characteristics: the consistency and the difference. Due to the unified multimodal annotation, existing methods are restricted in capturing differentiated information. However, additional uni-modal annotations are high time- and labor-cost. In this paper, we design a label generation module based on the self-supervised learning strategy to acquire independent unimodal supervisions. Then, joint training the multi-modal and uni-modal tasks to learn the consistency and difference, respectively. Moreover, during the training stage, we design a weight-adjustment strategy to balance the learning progress among different subtasks. That is to guide the subtasks to focus on samples with a larger difference between modality supervisions. Last, we conduct extensive experiments on three public multimodal baseline datasets. The experimental results validate the reliability and stability of auto-generated unimodal supervisions. On MOSI and MOSEI datasets, our method surpasses the current state-of-the-art methods. On the SIMS dataset, our method achieves comparable performance than humanannotated unimodal labels.","tags":[],"title":"Learning Modality-Specific Representations with Self-Supervised Multi-Task Learning for Multimodal Sentiment Analysis","type":"publication"},{"authors":["Kang Zhao","Hua Xu","Yue Cheng","Xiaoteng Li","Kai Gao"],"categories":[],"content":"","date":1612164244,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1612164244,"objectID":"77a9e3c7a954af1d8eeaab76f429bd74","permalink":"https://thu-xuhua.github.io/publication/rifre/","publishdate":"2021-03-04T15:24:04+08:00","relpermalink":"/publication/rifre/","section":"publication","summary":"Joint entity and relation extraction is an essential task in information extraction, which aims to extract all relational triples from unstructured text. However, few existing works consider possible relations information between entities before extracting them, which may lead to the fact that most of the extracted entities cannot constitute valid triples. In this paper, we propose a representation iterative fusion based on heterogeneous graph neural networks for relation extraction (RIFRE). We model relations and words as nodes on the graph and fuse the two types of semantic nodes by the message passing mechanism iteratively to obtain nodes representation that is more suitable for relation extraction tasks. The model performs relation extraction after nodes representation is updated. We evaluate RIFRE on two public relation extraction datasets: NYT and WebNLG. The results show that RIFRE can effectively extract triples and achieve state-of-the-art performance.1 Moreover, RIFRE is also suitable for the relation classification task, and significantly outperforms the previous methods on SemEval 2010 Task 8 datasets.","tags":[],"title":"Representation Iterative Fusion Based on Heterogeneous Graph Neural Network for Joint Entity and Relation Extraction","type":"publication"},{"authors":["KaichengYang","Hua Xu"],"categories":[],"content":"","date":1602491044,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1602491044,"objectID":"b35f6778f6190ac4e449c3199d0422ce","permalink":"https://thu-xuhua.github.io/publication/cmbert/","publishdate":"2020-09-01T16:24:04+08:00","relpermalink":"/publication/cmbert/","section":"publication","summary":"Multimodal sentiment analysis is an emerging research field that aims to enable machines to recognize, interpret, and express emotion. Through the cross-modal interaction, we can get more comprehensive emotional characteristics of the speaker. Bidirectional Encoder Representations from Transformers (BERT) is an efficient pre-trained language representation model. Fine-tuning it has obtained new state-of-the-art results on eleven natural language processing tasks like question answering and natural language inference. However, most previous works fine-tune BERT only base on text data, how to learn a better representation by introducing the multimodal information is still worth exploring. In this paper, we propose the Cross-Modal BERT (CM-BERT), which relies on the interaction of text and audio modality to fine-tune the pre-trained BERT model. As the core unit of the CM-BERT, masked multimodal attention is designed to dynamically adjust the weight of words by combining the information of text and audio modality. We evaluate our method on the public multimodal sentiment analysis datasets CMU-MOSI and CMU-MOSEI. The experiment results show that it has significantly improved the performance on all the metrics over previous baselines and text-only finetuning of BERT. Besides, we visualize the masked multimodal attention and proves that it can reasonably adjust the weight of words by introducing audio modality information.","tags":[],"title":"CM-BERT: Cross-Modal BERT for Text-Audio Sentiment Analysis","type":"publication"},{"authors":["Hua Xu","JiaLi"],"categories":[],"content":"","date":1598858644,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1598858644,"objectID":"ac11537788b44797f532b5c2b0333204","permalink":"https://thu-xuhua.github.io/publication/a-joint-model-of-extended-lda-and-ibtm-over-streaming-chinese-short-texts/","publishdate":"2020-08-31T15:24:04+08:00","relpermalink":"/publication/a-joint-model-of-extended-lda-and-ibtm-over-streaming-chinese-short-texts/","section":"publication","summary":"With the prevalent of short texts, discovering the topics within them has become an important task. Biterm Topic Model (BTM) is more suitable to discover topics on short texts than traditional topic models. However, there are still some challenges that dealing short texts with BTM will always ignore the document-topic semantic information and lack the true intentions of users. In addition, it is a static method and can not manage streaming short texts when a new one arrives immediately. In order to keep document-topic information and get the topic distribution of a new short text at once, we propose a joint model based on online algorithms of Latent Dirichlet Allocation (LDA) and BTM, which combines the merits of both models. Not only does it alleviate the sparsity when addressing short texts with the online algorithm of BTM, namely Incremental Biterm Topic Model (IBTM), but also keeps document-topic information with extended LDA. And considering the differences between English and Chinese text in writing, we use combined words in short texts as key words to extend the length of short texts and keep the true intensions of users. As shown in the experiment results on two real world datasets, our method is better than other baseline methods. In the end, we explain an application of our method in the task of discovering user interest tags.","tags":[],"title":"A joint model of extended LDA and IBTM over streaming Chinese short texts","type":"publication"},{"authors":["HuadongLi","Hua Xu"],"categories":[],"content":"","date":1598858644,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1598858644,"objectID":"65a5fe0a5ee8edf5aab919f2eac8fdfa","permalink":"https://thu-xuhua.github.io/publication/deep-reinforcement-learning-for-robust-emotional-classification-in-facial-expression-recognition/","publishdate":"2020-08-31T15:24:04+08:00","relpermalink":"/publication/deep-reinforcement-learning-for-robust-emotional-classification-in-facial-expression-recognition/","section":"publication","summary":"For emotion classification in facial expression recognition (FER), the performance of both traditional statistical methods and state-of-the-art deep learning methods are highly dependent on the quality of data. Traditional methods use image preprocessing (such as smoothing and segmentation) to improve image quality. However, the results still fail to meet the quality requirements of the emotion classifiers in FER. To address the above issues, this paper proposed a novel framework based on reinforcement learning for pre-selecting useful images(RLPS) for emotion classification in FER, which is made up of two modules: image selector and rough emotion classifier. Image selector is used to select useful images for emotion classification through reinforcement strategy and rough emotion classifier acts as a teacher to train image selector. Our framework improves classification performance by improving the quality of the dataset and can be applied to any classifier. Experiment results on RAF-DB, ExpW, and FER2013 datasets show that the proposed strategy achieves consistent improvements compared with the state-of-the-art emotion classification methods in FER","tags":[],"title":"Deep Reinforcement Learning for Robust Emotional Classification in Facial Expression Recognition","type":"publication"},{"authors":["XingweiHe","Hua Xu","JiaLi"],"categories":[],"content":"","date":1598858644,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1598858644,"objectID":"4fb03ebb5a5b433bffb675332469ded1","permalink":"https://thu-xuhua.github.io/publication/fastbtm-reducing-the-sampling-time-for-biterm-topic-model/","publishdate":"2020-08-31T15:24:04+08:00","relpermalink":"/publication/fastbtm-reducing-the-sampling-time-for-biterm-topic-model/","section":"publication","summary":"Due to the popularity of social networks, such as microblogs and Twitter, a vast amount of short text data is created every day. Much recent research in short text becomes increasingly significant, such as topic inference for short text. Biterm topic model (BTM) benefits from the word co-occurrence patterns of the corpus, which makes it perform better than conventional topic models in uncovering latent semantic relevance for short text. However, BTM resorts to Gibbs sampling to infer topics, which is very time consuming, especially for large-scale datasets or when the number of topics is extremely large. It requires O(K) operations per sample for K topics, where K denotes the number of topics in the corpus. In this paper, we propose an acceleration algorithm of BTM, FastBTM, using an efficient sampling method for BTM, which converges much faster than BTM without degrading topic quality. FastBTM is based on Metropolis-Hastings and alias method, both of which have been widely adopted in Latent Dirichlet Allocation (LDA) model and achieved outstanding speedup. Our FastBTM can effectively reduce the sampling complexity of biterm topic model from O(K) to O(1) amortized time. We carry out a number of experiments on three datasets including two short text datasets, Tweets2011 Collection dataset and Yahoo! Answers dataset, and one long document dataset, Enron dataset. Our experimental results show that when the number of topics K increases, the gap in running time speed between FastBTM and BTM gets especially larger. In addition, our FastBTM is effective for both short text datasets and long document datasets.","tags":[],"title":"FastBTM: Reducing the Sampling Time for Biterm Topic Model","type":"publication"},{"authors":["Hua Xu","JiyunZou"],"categories":[],"content":"","date":1598858644,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1598858644,"objectID":"88c292635c82e75a0782bf542017692d","permalink":"https://thu-xuhua.github.io/publication/hgfm/","publishdate":"2019-08-31T15:24:04+08:00","relpermalink":"/publication/hgfm/","section":"publication","summary":"To solve the problem of poor classification performance of multiple complex emotions in acoustic modalities, we propose a hierarchical grained and feature model (HGFM). The frame-level and utterance-level structures of acoustic samples are processed by the recurrent neural network. The model includes a frame-level representation module with before and after information, a utterance-level representation module with context information, and a different level acoustic feature fusion module. We take the output of frame-level structure as the input of utterance-level structure and extract the acoustic features of these two levels respectively for effective and complementary fusion. Experiments show that the proposed HGFM has better accuracy and robustness. By this method, we achieve the state-of-the-art performance on IEMOCAP and MELD datasets.","tags":[],"title":"HGFM : A Hierarchical Grained and Feature Model for Acoustic Emotion Recognition","type":"publication"},{"authors":["ZhijingWu","Hua Xu"],"categories":[],"content":"","date":1598858644,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1598858644,"objectID":"3167acc98ad4edf09bf7f9a7a2fe5049","permalink":"https://thu-xuhua.github.io/publication/improving-the-robustness-of-machine-reading-comprehension-model-with-hierarchical-knowledge-and-auxiliary-unanswerability-prediction/","publishdate":"2019-08-31T15:24:04+08:00","relpermalink":"/publication/improving-the-robustness-of-machine-reading-comprehension-model-with-hierarchical-knowledge-and-auxiliary-unanswerability-prediction/","section":"publication","summary":"Machine Reading Comprehension (MRC) aims to understand a passage and answer a series of related questions. With the development of deep learning and the release of large-scale MRC datasets, many end-to-end MRC neural networks have achieved remarkable success. However, these models are fragile and lack of robustness when there are some imperceptible adversarial perturbations in the input. In this paper, we propose an MRC model which has two main components to improve the robustness. On the one hand, we enhance the representation of the model by leveraging hierarchical knowledge from external knowledge bases. On the other hand, we introduce an auxiliary unanswerability prediction module and perform supervised multi-task learning along with a span prediction task. Experimental results on benchmark datasets show that our model can achieve consistent improvement compared with other strong baselines.","tags":[],"title":"Improving the Robustness of Machine Reading Comprehension Model with Hierarchical Knowledge and Auxiliary Unanswerability Prediction","type":"publication"},{"authors":["YuanYuan","Hua Xu"],"categories":[],"content":"","date":1598858644,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1598858644,"objectID":"ad2c2769a807e60939197f3659ff7519","permalink":"https://thu-xuhua.github.io/publication/objective-reduction-in-many-objective-optimization-evolutionary-multiobjective-approaches-and-comprehensive-analysis/","publishdate":"2020-08-31T15:24:04+08:00","relpermalink":"/publication/objective-reduction-in-many-objective-optimization-evolutionary-multiobjective-approaches-and-comprehensive-analysis/","section":"publication","summary":"Many-objective optimization problems bring great difficulties to the existing multiobjective evolutionary algorithms, in terms of selection operators, computational cost, visualization of the high-dimensional tradeoff front, and so on. Objective reduction can alleviate such difficulties by removing the redundant objectives in the original objective set, which has become one of the most important techniques in many-objective optimization. In this paper, we suggest to view objective reduction as a multiobjective search problem and introduce three multiobjective formulations of the problem, where the first two formulations are both based on preservation of the dominance structure and the third one utilizes the correlation between objectives. For each multiobjective formulation, a multiobjective objective reduction algorithm is proposed by employing the nondominated sorting genetic algorithm II to generate a Pareto front of nondominated objective subsets that can offer decision support to the user. Moreover, we conduct a comprehensive analysis of two major categories of objective reduction approaches based on several theorems, with the aim of revealing their strengths and limitations. Lastly, the performance of the proposed multiobjective algorithms is studied extensively on various benchmark problems and two real-world problems. Numerical results and comparisons are then shown to highlight the effectiveness and superiority of the proposed multiobjective algorithms over existing state-of-the-art approaches in the related field.","tags":[],"title":"Objective Reduction in Many-Objective Optimization: Evolutionary Multiobjective Approaches and Comprehensive Analysis","type":"publication"},{"authors":["WenmengYu","Hua Xu","FanyangMeng"],"categories":[],"content":"","date":1594538644,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1594538644,"objectID":"c8c941131fbf6a3d92f09b21b261513a","permalink":"https://thu-xuhua.github.io/publication/chsims/","publishdate":"2020-08-31T15:24:04+08:00","relpermalink":"/publication/chsims/","section":"publication","summary":"Previous studies in multimodal sentiment analysis have used limited datasets, which only contain unified multimodal annotations. However, the unified annotations do not always reflect the independent sentiment of single modalities and limit the model to capture the difference between modalities. In this paper, we introduce a Chinese single- and multi-modal sentiment analysis dataset, CH-SIMS, which contains 2,281 refined video segments in the wild with both multimodal and independent unimodal annotations. It allows researchers to study the interaction between modalities or use independent unimodal annotations for unimodal sentiment analysis.Furthermore, we propose a multi-task learning framework based on late fusion as the baseline. Extensive experiments on the CH-SIMS show that our methods achieve state-of-the-art performance and learn more distinctive unimodal representations. The full dataset and codes are available for use at https://github.com/thuiar/MMSA.","tags":[],"title":"CH-SIMS: A Chinese Multimodal Sentiment Analysis Dataset with Fine-grained Annotation of Modality","type":"publication"},{"authors":["YuxiangXie","Hua Xu","JiaoeLi","CongcongYang","KaiGao"],"categories":[],"content":"","date":1577836800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522715,"objectID":"d185324306c0bf4ed6fbfd5f3389cefe","permalink":"https://thu-xuhua.github.io/publication/heterogeneous-graph-neural-networks-for-noisy-few-shot-relation-classification/","publishdate":"2020-09-19T13:38:34.616343Z","relpermalink":"/publication/heterogeneous-graph-neural-networks-for-noisy-few-shot-relation-classification/","section":"publication","summary":"Relation classification is an essential and fundamental task in natural language processing. Distant supervised methods have achieved great success on relation classification, which improve the performance of the task through automatically extending the dataset. However, the distant supervised methods also bring the problem of wrong labeling. Inspired by people learning new knowledge from only a few samples, we focus on predicting formerly unseen classes with a few labeled data. In this paper, we propose a heterogeneous graph neural network for few-shot relation classification, which contains sentence nodes and entity nodes. We build the heterogeneous graph based on the message passing between entity nodes and sentence nodes in the graph, which can capture rich neighborhood information of the graph. Besides, we introduce adversarial learning for training a robust model and evaluate our heterogeneous graph neural networks under the scene of introducing different rates of noise data. Experimental results have demonstrated that our model outperforms the state-of-the-art baseline models on the FewRel dataset.","tags":["\"Relation extraction\"","\"Heterogeneous graph neural networks\"","\"Few-shot learning\"","\"Adversarial learning\""],"title":"Heterogeneous graph neural networks for noisy few-shot relation classification","type":"publication"},{"authors":["TingenLin","Hua Xu","HanleiZhang"],"categories":[],"content":"","date":1567236244,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1567236244,"objectID":"6805e8b3972313b1e33d1ada36cd3ade","permalink":"https://thu-xuhua.github.io/publication/discovering-new-intents-via-constrained-deep-adaptive-clustering-with-cluster-refinement/","publishdate":"2019-08-31T15:24:04+08:00","relpermalink":"/publication/discovering-new-intents-via-constrained-deep-adaptive-clustering-with-cluster-refinement/","section":"publication","summary":"Identifying new user intents is an essential task in the dialogue system. However, it is hard to get satisfying clustering results since the definition of intents is strongly guided by prior knowledge. Existing methods incorporate prior knowledge by intensive feature engineering, which not only leads to overfitting but also makes it sensitive to the number of clusters. In this paper, we propose constrained deep adaptive clustering with cluster refinement (CDAC+), an end-to-end clustering method that can naturally incorporate pairwise constraints as prior knowledge to guide the clustering process. Moreover, we refine the clusters by forcing the model to learn from the high confidence assignments. After eliminating low confidence assignments, our approach is surprisingly insensitive to the number of clusters. Experimental results on the three benchmark datasets show that our method can yield significant improvements over strong baselines.","tags":[],"title":"Discovering New Intents via Constrained Deep Adaptive Clustering with Cluster Refinement","type":"publication"},{"authors":["YuxiangXie","Hua Xu","CongcongYang"],"categories":[],"content":"","date":1567236244,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1567236244,"objectID":"d9772fcbba83c5a10407a3b1f722685d","permalink":"https://thu-xuhua.github.io/publication/multi-channel-convolutional-neural-networks-with-adversarial-training-for-few-shot-relation-classification/","publishdate":"2019-08-31T15:24:04+08:00","relpermalink":"/publication/multi-channel-convolutional-neural-networks-with-adversarial-training-for-few-shot-relation-classification/","section":"publication","summary":"The distant supervised (DS) method has improved the performance of relation classification (RC) by means of extending the dataset. However, DS also brings the problem of wrong labeling. Contrary to DS, the few-shot method relies on few supervised data to predict the unseen classes. In this paper, we use word embedding and position embedding to construct multi-channel vector representation and use the multi-channel convolutional method to extract features of sentences. Moreover, in order to alleviate few-shot learning to be sensitive to overfitting, we introduce adversarial learning for training a robust model. Experiments on the FewRel dataset show that our model achieves significant and consistent improvements on few-shot RC as compared with baselines.","tags":[],"title":"Multi-Channel Convolutional Neural Networks with Adversarial Training for Few-Shot Relation Classification","type":"publication"},{"authors":["TingenLin","Hua Xu"],"categories":[],"content":"","date":1561939200,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600519031,"objectID":"1984795fca17733159611b9de6ee29b1","permalink":"https://thu-xuhua.github.io/publication/deep-unknown-intent-detection-with-margin-loss/","publishdate":"2020-09-19T12:37:10.244831Z","relpermalink":"/publication/deep-unknown-intent-detection-with-margin-loss/","section":"publication","summary":"Identifying the unknown (novel) user intents that have never appeared in the training set is a challenging task in the dialogue system. In this paper, we present a two-stage method for detecting unknown intents. We use bidirectional long short-term memory (BiLSTM) network with the margin loss as the feature extractor. With margin loss, we can learn discriminative deep features by forcing the network to maximize inter-class variance and to minimize intra-class variance. Then, we feed the feature vectors to the density-based novelty detection algorithm, local outlier factor (LOF), to detect unknown intents. Experiments on two benchmark datasets show that our method can yield consistent improvements compared with the baseline methods.","tags":[],"title":"Deep Unknown Intent Detection with Margin Loss","type":"publication"},{"authors":["TingenLin","Hua Xu"],"categories":[],"content":"","date":1546300800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522716,"objectID":"a8639de67fa0ab3e0c38111ee425a908","permalink":"https://thu-xuhua.github.io/publication/a-post-processing-method-for-detecting-unknown-intent-of-dialogue-system-via-pre-trained-deep-neural-network-classifier/","publishdate":"2020-09-19T13:38:35.985341Z","relpermalink":"/publication/a-post-processing-method-for-detecting-unknown-intent-of-dialogue-system-via-pre-trained-deep-neural-network-classifier/","section":"publication","summary":"With the maturity and popularity of dialogue systems, detecting user’s unknown intent in dialogue systems has become an important task. It is also one of the most challenging tasks since we can hardly get examples, prior knowledge or the exact numbers of unknown intents. In this paper, we propose SofterMax and deep novelty detection (SMDN), a simple yet effective post-processing method for detecting unknown intent in dialogue systems based on pre-trained deep neural network classifiers. Our method can be flexibly applied on top of any classifiers trained in deep neural networks without changing the model architecture. We calibrate the confidence of the softmax outputs to compute the calibrated confidence score (i.e., SofterMax) and use it to calculate the decision boundary for unknown intent detection. Furthermore, we feed the feature representations learned by the deep neural networks into traditional novelty detection algorithm to detect unknown intents from different perspectives. Finally, we combine the methods above to perform the joint prediction. Our method classifies examples that differ from known intents as unknown and does not require any examples or prior knowledge of it. We have conducted extensive experiments on three benchmark dialogue datasets. The results show that our method can yield significant improvements compared with the state-of-the-art baselines1 1The code will be available at https://github.com/tnlin/SMDN..","tags":["\"Novelty detection\"","\"Open-world classification\"","\"Probability calibration\"","\"Platt scaling\"","\"Dialogue system\"","\"Deep neural network\""],"title":"A post-processing method for detecting unknown intent of dialogue system via pre-trained deep neural network classifier","type":"publication"},{"authors":["HongyanWang","Hua Xu","YuanYuan","XiaominSun","JunhuiDeng"],"categories":[],"content":"","date":1546300800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523728,"objectID":"7013dac9397a1bb754ec0da887b65a9d","permalink":"https://thu-xuhua.github.io/publication/balancing-exploration-and-exploitation-in-multiobjective-batch-bayesian-optimization/","publishdate":"2020-09-19T13:55:27.195975Z","relpermalink":"/publication/balancing-exploration-and-exploitation-in-multiobjective-batch-bayesian-optimization/","section":"publication","summary":"Many applications such as hyper-parameter tunning in Machine Learning can be casted to multiobjective black-box problems and it is challenging to optimize them. Bayesian Optimization (BO) is an effective method to deal with black-box functions. This paper mainly focuses on balancing exploration and exploitation in multi-objective black-box optimization problems by multiple samplings in BBO. In each iteration, multiple recommendations are generated via two different trade-off strategies respectively the expected improvement (EI) and a multiobjective framework with the mean and variance function of the GP posterior forming two conflict objectives. We compare our algorithm with ParEGO by running on 12 test functions. Hypervolume (HV, also known as S-metric) results show that our algorithm works well in exploration-exploitation trade-off for multiobjective black-box optimization problems.","tags":["\"batch bayesian optimization\"","\"expensive multiobjective optimization\"","\"exploration and exploitation\"","\"gaussian process\"","\"ParEGO\""],"title":"Balancing Exploration and Exploitation in Multiobjective Batch Bayesian Optimization","type":"publication"},{"authors":["HongyanWang","Hua Xu","YuanYuan","JunhuiDeng","XiaominSun"],"categories":[],"content":"","date":1546300800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523729,"objectID":"ef91c0c1dd573e34d7cc5ccf176eb998","permalink":"https://thu-xuhua.github.io/publication/noisy-multiobjective-black-box-optimization-using-bayesian-optimization/","publishdate":"2020-09-19T13:55:28.463251Z","relpermalink":"/publication/noisy-multiobjective-black-box-optimization-using-bayesian-optimization/","section":"publication","summary":"Expensive black-box problems are usually optimized by Bayesian Optimization (BO) since it can reduce evaluation costs via cheaper surrogates. The most popular model used in Bayesian Optimization is the Gaussian process (GP) whose posterior is based on a joint GP prior built by initial observations, so the posterior is also a Gaussian process. Observations are often not noise-free, so in most of these cases, a noisy transformation of the objective space is observed. Many single objective optimization algorithms have succeeded in extending efficient global optimization (EGO) to noisy circumstances, while ParEGO fails to consider noise. In order to deal with noisy expensive black-box problems, we extending ParEGO to noisy optimization according to adding a Gaussian noisy error while approximating the surrogate. We call it noisy-ParEGO and results of S-metric indicate that the algorithm works well on optimizing noisy expensive multiobjective black-box problems.","tags":["\"gaussian noise\"","\"gaussian process\"","\"ParEGO\"","\"black-box optimization\"","\"expensive multiobjective optimization\""],"title":"Noisy Multiobjective Black-Box Optimization Using Bayesian Optimization","type":"publication"},{"authors":["Hua Xu"],"categories":null,"content":"Course Classification: Public Elective Courses of Tsinghua University\nLecturer: Hua Xu\nTarget Audience: All Undergraduate Students\nTeaching Time：2019 - Today\n","date":1546272000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1546272000,"objectID":"407bfb7ec4580d302372641bb7e5039e","permalink":"https://thu-xuhua.github.io/talk/internet-product-design/","publishdate":"2019-01-01T00:00:00+08:00","relpermalink":"/talk/internet-product-design/","section":"talk","summary":"Public Elective Courses of Tsinghua University","tags":[],"title":"Internet Product Design","type":"talk"},{"authors":["KaiGao","Hua Xu","ChengliangGao","XiaominSun","JunhuiDeng","XiaomingZhang"],"categories":[],"content":"","date":1514764800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523732,"objectID":"0ccef8b0b6a04d04383a4ffb12728a87","permalink":"https://thu-xuhua.github.io/publication/two-stage-attention-network-for-aspect-level-sentiment-classification/","publishdate":"2020-09-19T13:55:31.487527Z","relpermalink":"/publication/two-stage-attention-network-for-aspect-level-sentiment-classification/","section":"publication","summary":"Currently, most of attention-based works adopt single-stage attention processes during generating context representations toward aspect, but their work lacks the deliberation process: A generated and aspect-related representation is directly used as final output without further polishing. In this work, we introduce the deliberation process to model context for further polishing of attention weights, and then propose a two-stage attention network for aspect-level sentiment classification. The network uses of a two-level attention model with LSTM, where the first-stage attention generates a raw aspect-related representation and the second-stage attention polishes and refines the raw representation by deliberation process. Since the deliberation component has global information what the representation to be generated might be, it has the potential to generate a better aspect-related representation by secondly looking into hidden state produced by LSTM. Experimental results on the dataset of SemEval-2016 task 5 about Laptop indicates that our model achieved the state-of-the-art accuracy of 76.56%.","tags":[],"title":"Two-Stage Attention Network for Aspect-Level Sentiment Classification","type":"publication"},{"authors":["JiaLi","Hua Xu"],"categories":[],"content":"","date":1472628244,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1472628244,"objectID":"bf0fc5a371becbc8bc558a236d65a68d","permalink":"https://thu-xuhua.github.io/publication/suggest-what-to-tag-recommending-more-precise-hashtags-based-on-users-dynamic-interests-and-streaming-tweet-content/","publishdate":"2020-08-31T15:24:04+08:00","relpermalink":"/publication/suggest-what-to-tag-recommending-more-precise-hashtags-based-on-users-dynamic-interests-and-streaming-tweet-content/","section":"publication","summary":"Twitter is an online social networking microblogging service that allows registered users to broadcast 140-character messages called tweets. The service has gained worldwide popularity since it was created in March 2006, with more than 316 million monthly active users in June 2015 who posted 500 million tweets per day. As the number of available tweets grows, the problem of managing tweets becomes extremely difficult, which could lead to information overload. To avoid this problem, people use the hashtag symbol # before a relevant keyword or phrase in their tweets to categorize those tweets and help them show more easily in each Twitter search. Furthermore, hashtags can be used to collect public opinions on events and their ideas at the individual, community or even the world level. Incorporating hashtags to obtain better performance such as sentiment classification and breaking events detection also has attracted considerable research attention in recent years. However, there are very few tweets containing hashtags, which impedes the quality of search results and their further usage in various applications. Therefore, hashtag recommendation has become a particularly important research problem. In this paper, we first propose a novel model, namely online Twitter-User LDA to learn Twitter users’ dynamic interests. Then considering the shortness, sparsity, and high volume of tweets, we introduce an effective method to discover the latent topics of streaming tweet content, which uses recently proposed incremental biterm topic model (IBTM). We finally design an automatic hashtag recommendation method called User-IBTM by combining the online Twitter-User LDA and IBTM. As shown in the experimental results on real world data from Twitter, our design method based on dynamic user interests and streaming tweet content significantly outperforms several other baseline methods and can suggest more precise hashtags.","tags":[],"title":"Suggest what to tag: Recommending more precise hashtags based on users’ dynamic interests and streaming tweet content","type":"publication"},{"authors":["WenhaoZhang","JianqiuJi","JunZhu","JianminLi","Hua Xu","BoZhang"],"categories":[],"content":"","date":1451606400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522718,"objectID":"ca2321ef330a70a051b1971c748ae7ae","permalink":"https://thu-xuhua.github.io/publication/bithash/","publishdate":"2020-09-19T13:38:37.383344Z","relpermalink":"/publication/bithash/","section":"publication","summary":"Locality Sensitive Hashing has been applied to detecting near-duplicate images, videos and web documents. In this paper we present a Bitwise Locality Sensitive method by using only one bit per hash value (BitHash), the storage space for storing hash values is significantly reduced, and the estimator can be computed much faster. The method provides an unbiased estimate of pairwise Jaccard similarity, and the estimator is a linear function of Hamming distance, which is very simple. We rigorously analyze the variance of One-Bit Min-Hash (BitHash), showing that for high Jaccard similarity. BitHash may provide accurate estimation, and as the pairwise Jaccard similarity increases, the variance ratio of BitHash over the original min-hash decreases. Furthermore, BitHash compresses each data sample into a compact binary hash code while preserving the pairwise similarity of the original data. The binary code can be used as a compressed and informative representation in replacement of the original data for subsequent processing. For example, it can be naturally integrated with a classifier like SVM. We apply BitHash to two typical applications, near-duplicate image detection and sentiment analysis. Experiments on real user’s photo collection and a popular sentiment analysis data set show that, the classification accuracy of our proposed method for two applications could approach the state-of-the-art method, while BitHash only requires a significantly smaller storage space.","tags":["\"Locality Sensitive Hashing\"","\"BitHash\"","\"Near-duplicate detection\"","\"Machine learning\"","\"Sentiment analysis\"","\"Storage efficiency\""],"title":"BitHash: An efficient bitwise Locality Sensitive Hashing method with applications","type":"publication"},{"authors":["YunfengXu","Hua Xu","DongwenZhang","YanZhang"],"categories":[],"content":"","date":1451606400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522717,"objectID":"495d62bb5e5488950d5cefcd567c9840","permalink":"https://thu-xuhua.github.io/publication/finding-overlapping-community-from-social-networks-based-on-community-forest-model/","publishdate":"2020-09-19T13:38:36.677344Z","relpermalink":"/publication/finding-overlapping-community-from-social-networks-based-on-community-forest-model/","section":"publication","summary":"Overlapping community detection is the key research work to discover and explore the social networks. A great deal of work has been devoted to detect overlapping communities, but no one can give a clear formula definition of community from the internal structure to the external boundary. More in depth, there are four challenges to existing research works. In this paper, firstly we propose overlapping community forest model and disjoint community forest model based on the community forest model, secondly give a clear formula definition of overlapping community and disjoint community based on the backbone degree and expansion, thirdly propose a novel algorithm to find overlapping communities based on the backbone degree and expansion to resolve the four challenges. This algorithm has better performance than four related algorithms mentioned by this paper in large scale social networks. It works well on American college football, Zachary’s Karate Club, Netscience-coauthor, Condensed matter collaborations, LFR etc. data sets.","tags":["\"Community detection\"","\"Social network\"","\"Expansion\"","\"Community forest model\""],"title":"Finding overlapping community from social networks based on community forest model","type":"publication"},{"authors":["JiaLi","Hua Xu"],"categories":[],"content":"","date":1451606400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523739,"objectID":"6e98a902ace34eeb8d2c7de9cf6425e5","permalink":"https://thu-xuhua.github.io/publication/user-ibtm/","publishdate":"2020-09-19T13:55:38.430568Z","relpermalink":"/publication/user-ibtm/","section":"publication","summary":"Twitter, the global social networking microblogging service, allows registered users to post 140-character messages known as tweets. People use the hashtag symbol `#' before a relevant keyword or phrase in their tweets to categorize the tweets and help them show more easily in a Twitter search. However, there are very few tweets contain hashtags, which impedes the quality of the search results and their applications. Therefore, how to automatically generate or recommend hashtags has become a particularly important academic research problem. Although many attempts have been made for solving this problem, previous methods mostly do not take the dynamic nature of hashtags into consideration. Furthermore, most previous work focuses on exploiting the similarity between tweets and ignores semantics in tweets.","tags":[],"title":"User-IBTM: An Online Framework for Hashtag Suggestion in Twitter","type":"publication"},{"authors":["Hua Xu"],"categories":null,"content":"Course Classification: Public Elective Courses of Tsinghua University\nLecturer: Hua Xu\nTarget Audience: All Undergraduate Students\nTeaching Time：2016 - Today\n","date":1451577600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1451577600,"objectID":"050eb509d73d451227a478cbd25c314e","permalink":"https://thu-xuhua.github.io/talk/intelligent-mobile-robot/","publishdate":"2016-01-01T00:00:00+08:00","relpermalink":"/talk/intelligent-mobile-robot/","section":"talk","summary":"Public Elective Courses of Tsinghua University","tags":[],"title":"Intelligent Mobile Robot: Design, Programming and Practice","type":"talk"},{"authors":["YunfengXu","Hua Xu","DongwenZhang"],"categories":[],"content":"","date":1420070400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522715,"objectID":"ca45fcc14ebd1b3801a6280c4cba52bc","permalink":"https://thu-xuhua.github.io/publication/a-novel-disjoint-community-detection-algorithm-for-social-networks-based-on-backbone-degree-and-expansion/","publishdate":"2020-09-19T13:38:35.312341Z","relpermalink":"/publication/a-novel-disjoint-community-detection-algorithm-for-social-networks-based-on-backbone-degree-and-expansion/","section":"publication","summary":"Community detection in social networks is a key point to discover the functions and structure of social networks. A great deal of work has been done for overlapping community detection and disjoint community detection, and numerous techniques such as spectral clustering, modularity maximization, random walks, differential equation, and statistical mechanics are used to identify a community in networks, but most of these work adopts pure mathematic and physical methods to discover communities from social networks, on the contrary ignoring the social and biological properties of communities and social networks. In this paper, firstly we propose the community forest model based on these social and biological properties to characterize the structure of real-world large-scale networks, secondly we mainly define a new metric named backbone degree to measure the strength of the edge and the similarity of vertices and give a new sense definition to community based on expansion, thirdly we develop a novel algorithm that based on backbone degree and expansion to discover disjoint communities from real social networks. This algorithm has better performance and effects compared with CNM and GN algorithms in computational cost and visibility. It has worked well on Email-Enron, American College Football, karate club etc. data sets.","tags":["\"Community detection\"","\"Social network\"","\"Expansion\"","\"Conductance\""],"title":"A novel disjoint community detection algorithm for social networks based on backbone degree and expansion","type":"publication"},{"authors":["KaiGao","Hua Xu","JiushuoWang"],"categories":[],"content":"","date":1420070400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522719,"objectID":"fb00db4f249eb0a0bb6dd341b1afe2e4","permalink":"https://thu-xuhua.github.io/publication/a-rule-based-approach-to-emotion-cause-detection-for-chinese-micro-blogs/","publishdate":"2020-09-19T13:38:38.726346Z","relpermalink":"/publication/a-rule-based-approach-to-emotion-cause-detection-for-chinese-micro-blogs/","section":"publication","summary":"Emotion analysis and emotion cause extraction are key research tasks in natural language processing and public opinion mining. This paper presents a rule-based approach to emotion cause component detection for Chinese micro-blogs. Our research has important scientific values on social network knowledge discovery and data mining. It also has a great potential in analyzing the psychological processes of consumers. Firstly, this paper proposes a rule-based system underlying the conditions that trigger emotions based on an emotional model. Secondly, this paper extracts the corresponding cause events in fine-grained emotions from the results of events, actions of agents and aspects of objects. Meanwhile, it is reasonable to get the proportions of different cause components under different emotions by constructing the emotional lexicon and identifying different linguistic features, and the proposed approach is based on Bayesian probability. Finally, this paper presents the experiments on an emotion corpus of Chinese micro-blogs. The experimental results validate the feasibility of the approach. The existing problems and the further works are also present at the end.","tags":["\"Text mining\"","\"Emotion causes\"","\"Micro-blog\"","\"Cause component proportion\""],"title":"A rule-based approach to emotion cause detection for Chinese micro-blogs","type":"publication"},{"authors":["YuanYuan","Hua Xu","BoWang"],"categories":[],"content":"","date":1420070400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523740,"objectID":"3959cdd0b5a4c70d47851688c5787682","permalink":"https://thu-xuhua.github.io/publication/an-experimental-investigation-of-variation-operators-in-reference-point-based-many-objective-optimization/","publishdate":"2020-09-19T13:55:39.432978Z","relpermalink":"/publication/an-experimental-investigation-of-variation-operators-in-reference-point-based-many-objective-optimization/","section":"publication","summary":"Reference-point based multi-objective evolutionary algorithms (MOEAs) have shown promising performance in many-objective optimization. However, most of existing research within this area focused on improving the environmental selection procedure, and little work has been done on the effect of variation operators. In this paper, we conduct an experimental investigation of variation operators in a typical reference-point based MOEA, i.e., NSGA-III. First, we provide a new NSGA-III variant, i.e., NSGA-III-DE, which introduces differential evolution (DE) operator into NSGA-III, and we further examine the effect of two main control parameters in NSGA-III-DE. Second, we have an experimental analysis of the search behavior of NSGA-III-DE and NSGA-III. We observe that NSGA-III-DE is generally better at exploration whereas NSGA-III normally has advantages in exploitation. Third, based on this observation, we present two other NSGA-III variants, where DE operator and genetic operators are simply combined to reproduce solutions. Experimental results on several benchmark problems show that very encouraging performance can be achieved by three suggested new NSGA-III variants. Our work also indicates that the performance of NSGA-III is significantly bottlenecked by its variation operators, providing opportunities for the study of the other alternative ones.","tags":["\"many-objective optimization\"","\"differential evolution\"","\"NSGA-III\"","\"variation operators\"","\"reference-point\""],"title":"An Experimental Investigation of Variation Operators in Reference-Point Based Many-Objective Optimization","type":"publication"},{"authors":["DongwenZhang","Hua Xu","ZengcaiSu","YunfengXu"],"categories":[],"content":"","date":1420070400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522720,"objectID":"40e454c53b8e10ba6e38793897e964c6","permalink":"https://thu-xuhua.github.io/publication/chinese-comments-sentiment-classification-based-on-word2vec-and-svmperf/","publishdate":"2020-09-19T13:38:40.089341Z","relpermalink":"/publication/chinese-comments-sentiment-classification-based-on-word2vec-and-svmperf/","section":"publication","summary":"Since the booming development of e-commerce in the last decade, the researchers have begun to pay more attention to extract the valuable information from consumers comments. Sentiment classification, which focuses on classify the comments into positive class and negative class according to the polarity of sentiment, is one of the studies. Machine learning-based method for sentiment classification becomes mainstream due to its outstanding performance. Most of the existing researches are centered on the extraction of lexical features and syntactic features, while the semantic relationships between words are ignored. In this paper, in order to get the semantic features, we propose a method for sentiment classification based on word2vec and SVMperf. Our research consists of two parts of work. First of all, we use word2vec to cluster the similar features for purpose of showing the capability of word2vec to capture the semantic features in selected domain and Chinese language. And then, we train and classify the comment texts using word2vec again and SVMperf. In the process, the lexicon-based and part-of-speech-based feature selection methods are respectively adopted to generate the training file. We conduct the experiments on the data set of Chinese comments on clothing products. The experimental results show the superior performance of our method in sentiment classification.","tags":["\"Sentiment classification\"","\"Word2vec\"","\"SVM\"","\"Semantic features\""],"title":"Chinese comments sentiment classification based on word2vec and SVMperf","type":"publication"},{"authors":["KaiGao","Hua Xu","JiushuoWang"],"categories":[],"content":"","date":1420070400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523741,"objectID":"eb8e5efc3b3554edc09e1c944ff88526","permalink":"https://thu-xuhua.github.io/publication/emotion-cause-detection-for-chinese-micro-blogs-based-on-ecocc-model/","publishdate":"2020-09-19T13:55:40.414417Z","relpermalink":"/publication/emotion-cause-detection-for-chinese-micro-blogs-based-on-ecocc-model/","section":"publication","summary":"Micro-blog emotion mining and emotion cause extraction are essential in social network data mining. This paper presents a novel approach on Chinese micro-blog emotion cause detection based on the ECOCC model, focusing on mining factors for eliciting some kinds of emotions. In order to do so, the corresponding emotion causes are extracted. Moreover, the proportions of different cause components under different emotions are also calculated by means of combining the emotional lexicon with multiple characteristics (e.g., emoticon, punctuation, etc.). Experimental results show the feasibility of the approach. The proposed approaches have important scientific values on social network knowledge discovery and data mining.","tags":[],"title":"Emotion Cause Detection for Chinese Micro-Blogs Based on ECOCC Model","type":"publication"},{"authors":["Hua Xu","WeiweiYang","JiushuoWang"],"categories":[],"content":"","date":1420070400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522718,"objectID":"aa88fc0a8b87d130c3a2b9ef4c196586","permalink":"https://thu-xuhua.github.io/publication/hierarchical-emotion-classification-and-emotion-component-analysis-on-chinese-micro-blog-posts/","publishdate":"2020-09-19T13:38:38.052343Z","relpermalink":"/publication/hierarchical-emotion-classification-and-emotion-component-analysis-on-chinese-micro-blog-posts/","section":"publication","summary":"Text emotion analysis has long been a hot topic. With the development of social network, text emotion analysis on micro-blog posts becomes a new trend in recent years. However, most researchers classify posts into coarse-grained emotion classes, which cannot depict the emotions accurately. Besides, flat classification is mostly adopted, which brings difficulty for classifiers when given a large dataset. In this paper, by data preprocessing, feature extraction and feature selection, we classify Chinese micro-blog posts into fine-grained emotion classes, employing hierarchical classification to improve the performance of classifiers. Moreover, based on the regression values in classification procedure, we propose an algorithm to detect the principal emotions in posts and calculate their ratios.","tags":["\"Micro-blog\"","\"Text mining\"","\"Emotion classification\"","\"Emotion component analysis\""],"title":"Hierarchical emotion classification and emotion component analysis on chinese micro-blog posts","type":"publication"},{"authors":["Hua Xu","FanZhang","WeiWang"],"categories":[],"content":"","date":1420070400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522720,"objectID":"f94f2e5520658436dee499d326958f94","permalink":"https://thu-xuhua.github.io/publication/implicit-feature-identification-in-chinese-reviews-using-explicit-topic-mining-model/","publishdate":"2020-09-19T13:38:39.422363Z","relpermalink":"/publication/implicit-feature-identification-in-chinese-reviews-using-explicit-topic-mining-model/","section":"publication","summary":"The essential work of feature-specific opinion mining is centered on the product features. Previous related research work has often taken into account explicit features but ignored implicit features, However, implicit feature identification, which can help us better understand the reviews, is an essential aspect of feature-specific opinion mining. This paper is mainly centered on implicit feature identification in Chinese product reviews. We think that based on the explicit synonymous feature group and the sentences which contain explicit features, several Support Vector Machine (SVM) classifiers can be established to classify the non-explicit sentences. Nevertheless, instead of simply using traditional feature selection methods, we believe an explicit topic model in which each topic is pre-defined could perform better. In this paper, we first extend a popular topic modeling method, called Latent Dirichlet Allocation (LDA), to construct an explicit topic model. Then some types of prior knowledge, such as: must-links, cannot-links and relevance-based prior knowledge, are extracted and incorporated into the explicit topic model automatically. Experiments show that the explicit topic model, which incorporates pre-existing knowledge, outperforms traditional feature selection methods and other existing methods by a large margin and the identification task can be completed better.","tags":["\"Opinion mining\"","\"Implicit feature\"","\"Topic model\"","\"Support vector machine\"","\"Product review\""],"title":"Implicit feature identification in Chinese reviews using explicit topic mining model","type":"publication"},{"authors":["BoWang","Hua Xu","YuanYuan"],"categories":[],"content":"","date":1388534400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523744,"objectID":"d7bba3e9273dc823226987afedbe2bb4","permalink":"https://thu-xuhua.github.io/publication/a-two-level-hierarchical-eda-using-conjugate-priori/","publishdate":"2020-09-19T13:55:43.415724Z","relpermalink":"/publication/a-two-level-hierarchical-eda-using-conjugate-priori/","section":"publication","summary":"Estimation of distribution algorithms (EDAs) are stochastic optimization methods that guide the search by building and sampling probabilistic models. Inspired by Bayesian inference, we proposed a two-level hierarchical model based on beta distribution. Beta distribution is the conjugate priori for binomial distribution. Besides, we introduced a learning rate function into the framework to control the model update. To demonstrate the effectiveness and applicability of our proposed algorithm, experiments are carried out on the 01-knapsack problems. Experimental results show that the proposed algorithm outperforms cGA, PBIL and QEA.","tags":["\"empirical study\"","\"combinatorial optimization\"","\"artificial intelligence\""],"title":"A Two-Level Hierarchical EDA Using Conjugate Priori","type":"publication"},{"authors":["YuanYuan","Hua Xu","BoWang"],"categories":[],"content":"","date":1388534400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523742,"objectID":"07fdb59a6a221f58189e33ee29a38145","permalink":"https://thu-xuhua.github.io/publication/an-improved-nsga-iii-procedure-for-evolutionary-many-objective-optimization/","publishdate":"2020-09-19T13:55:41.38986Z","relpermalink":"/publication/an-improved-nsga-iii-procedure-for-evolutionary-many-objective-optimization/","section":"publication","summary":"Many-objective (four or more objectives) optimization problems pose a great challenge to the classical Pareto-dominance based multi-objective evolutionary algorithms (MOEAs), such as NSGA-II and SPEA2. This is mainly due to the fact that the selection pressure based on Pareto-dominance degrades severely with the number of objectives increasing. Very recently, a reference-point based NSGA-II, referred as NSGA-III, is suggested to deal with many-objective problems, where the maintenance of diversity among population members is aided by supplying and adaptively updating a number of well-spread reference points. However, NSGA-III still relies on Pareto-dominance to push the population towards Pareto front (PF), leaving room for the improvement of its convergence ability. In this paper, an improved NSGA-III procedure, called 牟-NSGA-III, is proposed, aiming to better tradeoff the convergence and diversity in many-objective optimization. In 牟-NSGA-III, the non-dominated sorting scheme based on the proposed 牟-dominance is employed to rank solutions in the environmental selection phase, which ensures both convergence and diversity. Computational experiments have shown that 牟-NSGA-III is significantly better than the original NSGA-III and MOEA/D on most instances no matter in convergence and overall performance.","tags":["\"non-dominated sorting\"","\"NSGA-III\"","\"many-objective optimization\"","\"牟-dominance\""],"title":"An Improved NSGA-III Procedure for Evolutionary Many-Objective Optimization","type":"publication"},{"authors":["JingfeiDu","Hua Xu","XiaoqiuHuang"],"categories":[],"content":"","date":1388534400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522721,"objectID":"ac2e290c927f9283c96ae5146e882efc","permalink":"https://thu-xuhua.github.io/publication/box-office-prediction-based-on-microblog/","publishdate":"2020-09-19T13:38:40.789342Z","relpermalink":"/publication/box-office-prediction-based-on-microblog/","section":"publication","summary":"As the importance and popularity of online social media has become more obvious, there are more researches aiming at making use of information from them. One important topic of this is predicting the future with social media. This paper focuses on predicting box offices using microblog. Compared with previous work which makes use of the count of related microblogs simply, the information from social media has been utilized more deeply in this paper. Two sets of features have been extracted: count based features and content based features. For the former, the information in the aspect of users, which decrease the influence of garbage microblogs, has been exploited. For content based features, a new box office oriented semantic classification method has been provided to make the features more relative with box offices. Meanwhile, more complex machine learning models such as SVM and neutral network have been applied to the prediction method. Our prediction model is more accurate and reliable. With our prediction method, the data in Tencent microblog has been utilized to predict box offices of certain movies in China. With the results, the strength of our method and predictive power of online social media can be completely demonstrated.","tags":["\"Box office\"","\"Microblog\"","\"Social media\"","\"Prediction model\""],"title":"Box office prediction based on microblog","type":"publication"},{"authors":["WeiWang","Hua Xu","WeiweiYang","XiaoqiuHuang"],"categories":[],"content":"","date":1388534400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523745,"objectID":"82c559431cef96c2008a5dfa587e6094","permalink":"https://thu-xuhua.github.io/publication/constrained-hlda-for-topic-discovery-in-chinese-microblogs/","publishdate":"2020-09-19T13:55:44.415137Z","relpermalink":"/publication/constrained-hlda-for-topic-discovery-in-chinese-microblogs/","section":"publication","summary":"Since microblog service became information provider on web scale, research on microblog has begun to focus more on its content mining. Most research on microblog context is often based on topic models, such as: Latent Dirichlet Allocation(LDA) and its variations. However,there are some challenges in previous research. On one hand, the number of topics is fixed as a priori, but in real world, it is input by the users. On the other hand, it ignores the hierarchical information of topics and cannot grow structurally as more data are observed. In this paper, we propose a semi-supervised hierarchical topic model, which aims to explore more reasonable topics in the data space by incorporating some constraints into the modeling process that are extracted automatically. The new method is denoted as constrained hierarchical Latent Dirichlet Allocation (constrained-hLDA). We conduct experiments on Sina microblog, and evaluate the performance in terms of clustering and empirical likelihood. The experimental results show that constrained-hLDA has a significant improvement on the interpretability, and its predictive ability is also better than that of hLDA.","tags":[],"title":"Constrained-hLDA for Topic Discovery in Chinese Microblogs","type":"publication"},{"authors":["YuanYuan","Hua Xu","BoWang"],"categories":[],"content":"","date":1388534400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523743,"objectID":"01ba67e540e1e3d02a7bad897af1edad","permalink":"https://thu-xuhua.github.io/publication/evolutionary-many-objective-optimization-using-ensemble-fitness-ranking/","publishdate":"2020-09-19T13:55:42.386312Z","relpermalink":"/publication/evolutionary-many-objective-optimization-using-ensemble-fitness-ranking/","section":"publication","summary":"In this paper, a new framework, called ensemble fitness ranking (EFR), is proposed for evolutionary many-objective optimization that allows to work with different types of fitness functions and ensemble ranking schemes. The framework aims to rank the solutions in the population more appropriately by combing the ranking results from many simple individual rankers. As to the form of EFR, it can be regarded as an extension of average and maximum ranking methods which have been shown promising for many-objective problems. The significant change is that EFR adopts more general fitness functions instead of objective functions, which would make it easier for EFR to balance the convergence and diversity in many-objective optimization. In the experimental studies, the influence of several fitness functions and ensemble ranking schemes on the performance of EFR is fist investigated. Afterwards, EFR is compared with two state-of-the-art methods (MOEA/D and NSGA-III) on well-known test problems. The computational results show that EFR significantly outperforms MOEA/D and NSGA-III on most instances, especially for those having a high number of objectives.","tags":["\"average ranking\"","\"MOEA/D\"","\"maximum ranking\"","\"ensemble fitness ranking\"","\"fitness function\"","\"NSGA-III\"","\"many-objective optimization\""],"title":"Evolutionary Many-Objective Optimization Using Ensemble Fitness Ranking","type":"publication"},{"authors":["WeiyuanLi","Hua Xu"],"categories":[],"content":"","date":1388534400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522722,"objectID":"eb4f236771b0188dbca7d3092873ab5e","permalink":"https://thu-xuhua.github.io/publication/text-based-emotion-classification-using-emotion-cause-extraction/","publishdate":"2020-09-19T13:38:41.466347Z","relpermalink":"/publication/text-based-emotion-classification-using-emotion-cause-extraction/","section":"publication","summary":"In recent years, increasing impact of social networks on people’s opinions and decision making has attracted lots of attention. Microblogging, one of the most popular social network applications that allows people to share ideas and discuss over various topics, is taken as a rich resource of opinion and emotion data. In this paper, we propose and implement a novel method for identifying emotions in microblog posts. Unlike traditional approaches which are mostly based on statistical methods, we try to infer and extract the reasons of emotions by importing knowledge and theories from other fields such as Sociology. Based on the theory that a triggering cause event is an integral part of emotion, the technique of emotion cause extraction is used as a crucial step to improve the quality of selected features. First, after thorough analysis on sample data we constructed an automatic rule-based system to detect and extract the cause event of each emotional post. We build an emotion corpus with Chinese microblog posts labeled by human annotators. Then a classifier is trained to classify emotions in microblog posts based on extracted cause events. The overall performance of our system is very promising. The experiment results show that our approach is effective in selecting informative features. Our system outperformed the baseline noticeably in most cases, suggesting its great potential. This exploration should provide a new way to look at the emotion classification task and lay the ground for future research on textual emotion processing.","tags":["\"Emotion classification\"","\"Emotion cause extraction\"","\"Microblogging\"","\"Weibo\""],"title":"Text-based emotion classification using emotion cause extraction","type":"publication"},{"authors":["WeiWang","Hua Xu","Xiaoqiu Huang"],"categories":[],"content":"","date":1380585600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523746,"objectID":"19ba18be40238c84b2af736b7ef526bc","permalink":"https://thu-xuhua.github.io/publication/implicit-feature-detection-via-a-constrained-topic-model-and-svm/","publishdate":"2020-09-19T13:55:45.407589Z","relpermalink":"/publication/implicit-feature-detection-via-a-constrained-topic-model-and-svm/","section":"publication","summary":"","tags":[],"title":"Implicit Feature Detection via a Constrained Topic Model and SVM","type":"publication"},{"authors":["YuanYuan","Hua Xu","JiadongYang"],"categories":[],"content":"","date":1356998400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522723,"objectID":"46c5262473bbe0d43c6d815a83c10361","permalink":"https://thu-xuhua.github.io/publication/a-hybrid-harmony-search-algorithm-for-the-flexible-job-shop-scheduling-problem/","publishdate":"2020-09-19T13:38:42.870343Z","relpermalink":"/publication/a-hybrid-harmony-search-algorithm-for-the-flexible-job-shop-scheduling-problem/","section":"publication","summary":"In this paper, a novel hybrid harmony search (HHS) algorithm based on the integrated approach, is proposed for solving the flexible job shop scheduling problem (FJSP) with the criterion to minimize makespan. First of all, to make the harmony search (HS) algorithm adaptive to the FJSP, the converting techniques are developed to convert the continuous harmony vector to a kind of discrete two-vector code for the FJSP. Secondly, the harmony vector is mapped into a feasible active schedule through effectively decoding the transformed two-vector code, which could largely reduce the search space. Thirdly, a resultful initialization scheme combining heuristic and random strategies is introduced to make the initial harmony memory (HM) occur with certain quality and diversity. Furthermore, a local search procedure is embedded in the HS algorithm to enhance the local exploitation ability, whereas HS is employed to perform exploration by evolving harmony vectors in the HM. To speed up the local search process, the improved neighborhood structure based on common critical operations is presented in detail. Empirical results on various benchmark instances validate the effectiveness and efficiency of our proposed algorithm. Our work also indicates that a well designed HS-based method is a competitive alternative for addressing the FJSP.","tags":["\"Scheduling\"","\"Flexible job shop\"","\"Harmony search\"","\"Local search\"","\"Neighborhood structure\"","\"Makespan\""],"title":"A hybrid harmony search algorithm for the flexible job shop scheduling problem","type":"publication"},{"authors":["YuanYuan","Hua Xu"],"categories":[],"content":"","date":1356998400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523747,"objectID":"33634febc225010f12f0381325109774","permalink":"https://thu-xuhua.github.io/publication/a-memetic-algorithm-for-the-multi-objective-flexible-job-shop-scheduling-problem/","publishdate":"2020-09-19T13:55:46.409Z","relpermalink":"/publication/a-memetic-algorithm-for-the-multi-objective-flexible-job-shop-scheduling-problem/","section":"publication","summary":"In this paper, a new memetic algorithm (MA) is proposed for the muti-objective flexible job shop scheduling problem (MO-FJSP) with the objectives to minimize the makespan, total workload and critical workload. By using well-designed chromosome encoding/decoding scheme and genetic operators, the non-dominated sorting genetic algorithm II (NSGA-II) is first adapted for the MO-FJSP. Then the MA is developed by incorporating a novel local search algorithm into the adapted NSGA-II, where several mechanisms to balance the genetic search and local search are employed. In the proposed local search, a hierarchical strategy is adopted to handle the three objectives, which mainly considers the minimization of makespan, while the concern of the other two objectives is reflected in the order of trying all the possible actions that could generate the acceptable neighbor. Experimental results on well-known benchmark instances show that the proposed MA outperforms significantly two off-the-shelf multi-objective evolutionary algorithms and four state-of-the-art algorithms specially proposed for the MO-FJSP.","tags":["\"local search\"","\"non-dominated sorting genetic algorithm ii (nsga-ii)\"","\"muti-objective\"","\"flexible job shop scheduling\"","\"memetic algorithm\""],"title":"A Memetic Algorithm for the Multi-Objective Flexible Job Shop Scheduling Problem","type":"publication"},{"authors":["YuanYuan","Hua Xu"],"categories":[],"content":"","date":1356998400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522722,"objectID":"f70bae8e147e1bf4c12326da9d9623ba","permalink":"https://thu-xuhua.github.io/publication/an-integrated-search-heuristic-for-large-scale-flexible-job-shop-scheduling-problems/","publishdate":"2020-09-19T13:38:42.161342Z","relpermalink":"/publication/an-integrated-search-heuristic-for-large-scale-flexible-job-shop-scheduling-problems/","section":"publication","summary":"The flexible job shop scheduling problem (FJSP) is a generalization of the classical job shop scheduling problem (JSP), where each operation is allowed to be processed by any machine from a given set, rather than one specified machine. In this paper, two algorithm modules, namely hybrid harmony search (HHS) and large neighborhood search (LNS), are developed for the FJSP with makespan criterion. The HHS is an evolutionary-based algorithm with the memetic paradigm, while the LNS is typical of constraint-based approaches. To form a stronger search mechanism, an integrated search heuristic, denoted as HHS/LNS, is proposed for the FJSP based on the two algorithms, which starts with the HHS, and then the solution is further improved by the LNS. Computational simulations and comparisons demonstrate that the proposed HHS/LNS shows competitive performance with state-of-the-art algorithms on large-scale FJSP problems, and some new upper bounds among the unsolved benchmark instances have even been found.","tags":["\"Scheduling\"","\"Flexible job shop\"","\"Harmony search\"","\"Large neighborhood search\"","\"Makespan\""],"title":"An integrated search heuristic for large-scale flexible job shop scheduling problems","type":"publication"},{"authors":["JiadongYang","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1356998400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522724,"objectID":"e0c67343da5590e9a45a78068d487e4e","permalink":"https://thu-xuhua.github.io/publication/effective-search-for-genetic-based-machine-learning-systems-via-estimation-of-distribution-algorithms-and-embedded-feature-reduction-techniques/","publishdate":"2020-09-19T13:38:43.598345Z","relpermalink":"/publication/effective-search-for-genetic-based-machine-learning-systems-via-estimation-of-distribution-algorithms-and-embedded-feature-reduction-techniques/","section":"publication","summary":"Genetic-based machine learning (GBML) systems, which employ evolutionary algorithms (EAs) as search mechanisms, evolve rule-based classification models to represent target concepts. Compared to Michigan-style GBML, Pittsburgh-style GBML is expected to achieve more compact solutions. It has been shown that standard recombination operators in EAs do not assure an effective evolutionary search to solve sophisticated problems that contain strong interactions between features. On the other hand, when dealing with real-world classification tasks, irrelevant features not only complicate the problem but also incur unnecessary matchings in GBML systems, which increase the computational cost a lot. To handle the two problems mentioned above in an integrated manner, a new Pittsburgh-style GBML system is proposed. In the proposed method, classifiers are generated and recombined at two levels. At the high level, classifiers are recombined by rule-wise uniform crossover operators since each classifier consists of a variable-size rule set. At the low level, single rules contained in classifiers are reproduced via sampling Bayesian networks that characterize the global statistical information extracted from promising rules found so far. Furthermore, according to the statistical information in the rule population, an embedded approach is presented to detect and remove redundant features incrementally following the evolution of rule population. Results of empirical evaluation show that the proposed method outperforms the original Pittsburgh-style GBML system in terms of classification accuracy while reducing the computational cost. Furthermore, the proposed method is also competitive to other non-evolutionary, highly used machine learning methods. With respect to the performance of feature reduction, the proposed embedded approach is able to deliver solutions with higher classification accuracy when removing the same number of features as other feature reduction techniques do.","tags":["\"Genetic-based machine learning systems\"","\"Estimation of distribution algorithms\"","\"Features reduction\"","\"Evolutionary computation\""],"title":"Effective search for genetic-based machine learning systems via estimation of distribution algorithms and embedded feature reduction techniques","type":"publication"},{"authors":["YuanYuan","Hua Xu"],"categories":[],"content":"","date":1356998400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522724,"objectID":"0d4a7810a06e4b6916c5f71d8b49e6ed","permalink":"https://thu-xuhua.github.io/publication/flexible-job-shop-scheduling-using-hybrid-differential-evolution-algorithms/","publishdate":"2020-09-19T13:38:44.301342Z","relpermalink":"/publication/flexible-job-shop-scheduling-using-hybrid-differential-evolution-algorithms/","section":"publication","summary":"This paper proposes hybrid differential evolution (HDE) algorithms for solving the flexible job shop scheduling problem (FJSP) with the criterion to minimize the makespan. Firstly, a novel conversion mechanism is developed to make the differential evolution (DE) algorithm that works on the continuous domain adaptive to explore the problem space of the discrete FJSP. Secondly, a local search algorithm based on the critical path is embedded in the DE framework to balance the exploration and exploitation by enhancing the local searching ability. In addition, in the local search phase, the speed-up method to find an acceptable schedule within the neighborhood structure is presented to improve the efficiency of whole algorithms. Extensive computational results and comparisons show that the proposed algorithms are very competitive with the state of the art, some new best known solutions for well known benchmark instances have even been found.","tags":["\"Scheduling\"","\"Flexible job shop\"","\"Differential evolution\"","\"Local search\"","\"Neighborhood structure\"","\"Makespan\""],"title":"Flexible job shop scheduling using hybrid differential evolution algorithms","type":"publication"},{"authors":["WeiWang","Hua Xu","WeiWan"],"categories":[],"content":"","date":1356998400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522725,"objectID":"ff12f93c98434c71db85027a1a96d805","permalink":"https://thu-xuhua.github.io/publication/implicit-feature-identification-via-hybrid-association-rule-mining/","publishdate":"2020-09-19T13:38:44.974341Z","relpermalink":"/publication/implicit-feature-identification-via-hybrid-association-rule-mining/","section":"publication","summary":"In sentiment analysis, a finer-grained opinion mining method not only focuses on the view of the product itself, but also focuses on product features, which can be a component or attribute of the product. Previous related research mainly relied on explicit features but ignored implicit features. However, the implicit features, which are implied by some words or phrases, are so significant that they can express the users’ opinion and help us to better understand the users’ comments. It is a big challenge to detect these implicit features in Chinese product reviews, due to the complexity of Chinese. This paper is mainly centered on implicit features identification in Chinese product reviews. A novel hybrid association rule mining method is proposed for this task. The core idea of this approach is mining as many association rules as possible via several complementary algorithms. Firstly, we extract candidate feature indicators based word segmentation, part-of-speech (POS) tagging and feature clustering, then compute the co-occurrence degree between the candidate feature indicators and the feature words using five collocation extraction algorithms. Each indicator and the corresponding feature word constitute a rule (feature indicator → feature word). The best rules in five different rule sets are chosen as the basic rules. Next, three methods are proposed to mine some possible reasonable rules from the lower co-occurrence feature indicators and non indicator words. Finally, the latest rules are used to identify implicit features and the results are compared with the previous. Experiment results demonstrate that our proposed approach is competent at the task, especially via using several expanding methods. The recall is effectively improved, suggesting that the shortcomings of the basic rules have been overcome to certain extent. Besides those high co-occurrence degree indicators, the final rules also contain uncommon rules.","tags":["\"Opinion mining\"","\"Implicit features\"","\"Hybrid association rule mining\"","\"Collocation extraction\""],"title":"Implicit feature identification via hybrid association rule mining","type":"publication"},{"authors":["Hua Xu"],"categories":null,"content":"Course Classification: Public Elective Courses of Tsinghua University\nLecturer: Hua Xu\nTarget Audience: All Graduate and Undergraduate Students\nTeaching Time：2013 - Today\n","date":1356969600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1356969600,"objectID":"e7d4ecfaf4b1f29cd65dd9ec01f1f6ad","permalink":"https://thu-xuhua.github.io/talk/industrial-data-mining/","publishdate":"2013-01-01T00:00:00+08:00","relpermalink":"/talk/industrial-data-mining/","section":"talk","summary":"Public Elective Courses of Tsinghua University","tags":[],"title":"Industrial Data Mining","type":"talk"},{"authors":["JiadongYang","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1325376000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522727,"objectID":"f75a908859cf123bb5152d3a7c4fe155","permalink":"https://thu-xuhua.github.io/publication/effective-search-for-pittsburgh-learning-classifier-systems-via-estimation-of-distribution-algorithms/","publishdate":"2020-09-19T13:38:46.367342Z","relpermalink":"/publication/effective-search-for-pittsburgh-learning-classifier-systems-via-estimation-of-distribution-algorithms/","section":"publication","summary":"Pittsburgh-style learning classifier systems (LCSs), in which an entire candidate solution is represented as a set of variable number of rules, combine supervised learning with genetic algorithms (GAs) to evolve rule-based classification models. It has been shown that standard crossover operators in GAs do not guarantee an effective evolutionary search in many sophisticated problems that contain strong interactions between features. In this paper, we propose a Pittsburgh-style learning classifier system based on the Bayesian optimization algorithm with the aim of improving the effectiveness and efficiency of the rule structure exploration. In the proposed method, classifiers are generated and recombined at two levels. At the lower level, single rules contained in classifiers are produced by sampling Bayesian networks which characterize the global statistical information extracted from the current promising rules in the search space. At the higher level, classifiers are recombined by rule-wise uniform crossover operators to keep the semantics of rules in each classifier. Experimental studies on both artificial and real world binary classification problems show that the proposed method converges faster while achieving solutions with the same or even higher accuracy compared with the original Pittsburgh-style LCSs.","tags":["\"Learning classifier system\"","\"Genetics-based machine learning\"","\"Estimation of distribution algorithm\"","\"Bayesian optimization algorithm\"","\"Evolutionary computation\""],"title":"Effective search for Pittsburgh learning classifier systems via estimation of distribution algorithms","type":"publication"},{"authors":["WenhaoZhang","Hua Xu","WeiWan"],"categories":[],"content":"","date":1325376000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522726,"objectID":"3491f2d19347a554bb739ade0035aded","permalink":"https://thu-xuhua.github.io/publication/weakness-finder/","publishdate":"2020-09-19T13:38:45.661348Z","relpermalink":"/publication/weakness-finder/","section":"publication","summary":"Finding the weakness of the products from the customers’ feedback can help manufacturers improve their product quality and competitive strength. In recent years, more and more people express their opinions about products online, and both the feedback of manufacturers’ products or their competitors’ products could be easily collected. However, it’s impossible for manufacturers to read every review to analyze the weakness of their products. Therefore, finding product weakness from online reviews becomes a meaningful work. In this paper, we introduce such an expert system, Weakness Finder, which can help manufacturers find their product weakness from Chinese reviews by using aspects based sentiment analysis. An aspect is an attribute or component of a product, such as price, degerm, moisturizing are the aspects of the body wash products. Weakness Finder extracts the features and groups explicit features by using morpheme based method and Hownet based similarity measure, and identify and group the implicit features with collocation selection method for each aspect. Then utilize sentence based sentiment analysis method to determine the polarity of each aspect in sentences. The weakness of product could be found because the weakness is probably the most unsatisfied aspect in customers’ reviews, or the aspect which is more unsatisfied when compared with their competitor’s product reviews. Weakness Finder has been used to help a body wash manufacturer find their product weakness, and our experimental results demonstrate the good performance of the Weakness Finder.","tags":["\"Product weakness\"","\"Business intelligence\"","\"Sentiment analysis\"","\"Feature grouping\""],"title":"Weakness Finder: Find product weakness from Chinese reviews by using aspects based sentiment analysis","type":"publication"},{"authors":["YunWen","Hua Xu","JiadongYang"],"categories":[],"content":"","date":1293840000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522729,"objectID":"99cf60281c3f77112b52eefd34837b20","permalink":"https://thu-xuhua.github.io/publication/a-heuristic-based-hybrid-genetic-variable-neighborhood-search-algorithm-for-task-scheduling-in-heterogeneous-multiprocessor-system/","publishdate":"2020-09-19T13:38:49.188341Z","relpermalink":"/publication/a-heuristic-based-hybrid-genetic-variable-neighborhood-search-algorithm-for-task-scheduling-in-heterogeneous-multiprocessor-system/","section":"publication","summary":"Effective task scheduling, which is essential for achieving high performance in a heterogeneous multiprocessor system, remains a challenging problem despite extensive studies. In this article, a heuristic-based hybrid genetic-variable neighborhood search algorithm is proposed for the minimization of makespan in the heterogeneous multiprocessor scheduling problem. The proposed algorithm distinguishes itself from many existing genetic algorithm (GA) approaches in three aspects. First, it incorporates GA with the variable neighborhood search (VNS) algorithm, a local search metaheuristic, to exploit the intrinsic structure of the solutions for guiding the exploration process of GA. Second, two novel neighborhood structures are proposed, in which problem-specific knowledge concerned with load balancing and communication reduction is utilized respectively, to improve both the search quality and efficiency of VNS. Third, the proposed algorithm restricts the use of GA to evolve the task-processor mapping solutions, while taking advantage of an upward-ranking heuristic mostly used by traditional list scheduling approaches to determine the task sequence assignment in each processor. Empirical results on benchmark task graphs of several well-known parallel applications, which have been validated by the use of non-parametric statistical tests, show that the proposed algorithm significantly outperforms several related algorithms in terms of the schedule quality. Further experiments are carried out to reveal that the proposed algorithm is able to maintain high performance within a wide range of parameter settings.","tags":["\"Variable neighborhood search\"","\"Genetic algorithm\"","\"Hybrid metaheuristic\"","\"Memetic algorithm\"","\"Heterogeneous multiprocessor scheduling\""],"title":"A heuristic-based hybrid genetic-variable neighborhood search algorithm for task scheduling in heterogeneous multiprocessor system","type":"publication"},{"authors":["AnqiCui","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1293840000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522728,"objectID":"4889a4457c74624c756d2a8d8d4bb58a","permalink":"https://thu-xuhua.github.io/publication/an-elman-neural-network-based-model-for-predicting-anti-germ-performances-and-ingredient-levels-with-limited-experimental-data/","publishdate":"2020-09-19T13:38:47.83434Z","relpermalink":"/publication/an-elman-neural-network-based-model-for-predicting-anti-germ-performances-and-ingredient-levels-with-limited-experimental-data/","section":"publication","summary":"Anti-germ performance test is critical in the production of detergents. However, actual biochemical tests are often costly and time-consuming. In this paper, we present an Elman neural network-based model to predict detergents’ anti-germ performance and ingredient levels, respectively. The model made it much faster and cost effective than doing actual biochemical tests. We also present preprocessing methods that can reduce data conflicts while keeping the monotonicity on limited experimental data. The model can find out the relationship between ingredient levels and the corresponding anti-germ performance, which is not widely used in solving biochemical problems. The results of experiments are generated on the base of two detergent products for two types of bacteria, and appear reasonable according to natural rules. The prediction results show a high accuracy and fitting with the monotonicity rule mostly.","tags":["\"Anti-germ performance prediction\"","\"Ingredient level prediction\"","\"Artificial neural networks\"","\"Monotonicity rule\"","\"Preprocessing methods\""],"title":"An Elman neural network-based model for predicting anti-germ performances and ingredient levels with limited experimental data","type":"publication"},{"authors":["ZhongwuZhai","BingLiu","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1293840000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523750,"objectID":"0b6bbf66b6b5dc4e868add1d7a4780d6","permalink":"https://thu-xuhua.github.io/publication/clustering-product-features-for-opinion-mining/","publishdate":"2020-09-19T13:55:49.375311Z","relpermalink":"/publication/clustering-product-features-for-opinion-mining/","section":"publication","summary":"In sentiment analysis of product reviews, one important problem is to produce a summary of opinions based on product features/attributes (also called aspects). However, for the same feature, people can express it with many different words or phrases. To produce a useful summary, these words and phrases, which are domain synonyms, need to be grouped under the same feature group. Although several methods have been proposed to extract product features from reviews, limited work has been done on clustering or grouping of synonym features. This paper focuses on this task. Classic methods for solving this problem are based on unsupervised learning using some forms of distributional similarity. However, we found that these methods do not do well. We then model it as a semi-supervised learning problem. Lexical characteristics of the problem are exploited to automatically identify some labeled examples. Empirical evaluation shows that the proposed method outperforms existing state-of-the-art methods by a large margin.","tags":["\"product feature grouping\"","\"opinion mining\""],"title":"Clustering Product Features for Opinion Mining","type":"publication"},{"authors":["ZhongwuZhai","BingLiu","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1293840000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523749,"objectID":"8045747747311ac63a85201fbab02533","permalink":"https://thu-xuhua.github.io/publication/constrained-lda-for-grouping-product-features-in-opinion-mining/","publishdate":"2020-09-19T13:55:48.426867Z","relpermalink":"/publication/constrained-lda-for-grouping-product-features-in-opinion-mining/","section":"publication","summary":"In opinion mining of product reviews, one often wants to produce a summary of opinions based on product features. However, for the same feature, people can express it with different words and phrases. To produce an effective summary, these words and phrases, which are domain synonyms, need to be grouped under the same feature. Topic modeling is a suitable method for the task. However, instead of simply letting topic modeling find groupings freely, we believe it is possible to do better by giving it some pre-existing knowledge in the form of automatically extracted constraints. In this paper, we first extend a popular topic modeling method, called Latent Dirichlet Allocation (LDA), with the ability to process large scale constraints. Then, two novel methods are proposed to extract two types of constraints automatically. Finally, the resulting constrained-LDA and the extracted constraints are applied to group product features. Experiments show that constrained-LDA outperforms the original LDA and the latest mLSA by a large margin.","tags":[],"title":"Constrained LDA for Grouping Product Features in Opinion Mining","type":"publication"},{"authors":["ZhongwuZhai","Hua Xu","BadaKang","PeifaJia"],"categories":[],"content":"","date":1293840000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522730,"objectID":"d36c5e5a1ea50571022a176b086059bf","permalink":"https://thu-xuhua.github.io/publication/exploiting-effective-features-for-chinese-sentiment-classification/","publishdate":"2020-09-19T13:38:49.919342Z","relpermalink":"/publication/exploiting-effective-features-for-chinese-sentiment-classification/","section":"publication","summary":"Features play a fundamental role in sentiment classification. How to effectively select different types of features to improve sentiment classification performance is the primary topic of this paper. Ngram features are commonly employed in text classification tasks; in this paper, sentiment-words, substrings, substring-groups, and key-substring-groups, which have never been considered in sentiment classification area before, are also extracted as features. The extracted features are then compared and analyzed. To demonstrate generality, we use two authoritative Chinese data sets in different domains to conduct our experiments. Our statistical analysis of the experimental results indicate the following: (1) different types of features possess different discriminative capabilities in Chinese sentiment classification; (2) character bigram features perform the best among the Ngram features; (3) substring-group features have greater potential to improve the performance of sentiment classification by combining substrings of different lengths; (4) sentiment words or phrases extracted from existing sentiment lexicons are not effective for sentiment classification; (5) effective features are usually at varying lengths rather than fixed lengths.","tags":["\"Sentiment classification\"","\"Substring features\"","\"Substring-group\"","\"Suffix tree\""],"title":"Exploiting effective features for Chinese sentiment classification","type":"publication"},{"authors":["ZhongwuZhai","BingLiu","LeiZhang","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1293840000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523748,"objectID":"dc94ebc0f2c13a4d42abbfe0640ccf11","permalink":"https://thu-xuhua.github.io/publication/identifying-evaluative-sentences-in-online-discussions/","publishdate":"2020-09-19T13:55:47.457403Z","relpermalink":"/publication/identifying-evaluative-sentences-in-online-discussions/","section":"publication","summary":"","tags":[],"title":"Identifying Evaluative Sentences in Online Discussions","type":"publication"},{"authors":["WeiWan","Hua Xu","WenhaoZhang","XinchengHu","GangDeng"],"categories":[],"content":"","date":1293840000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522727,"objectID":"6a0bd186488f3ae758f71e2d65b97a67","permalink":"https://thu-xuhua.github.io/publication/questionnaires-based-skin-attribute-prediction-using-elman-neural-network/","publishdate":"2020-09-19T13:38:47.128344Z","relpermalink":"/publication/questionnaires-based-skin-attribute-prediction-using-elman-neural-network/","section":"publication","summary":"Skin attribute tests, especially for women, have become critical in the development of daily cosmetics in recent years. However, clinical skin attribute testing is often costly and time consuming. In this paper, a novel prediction approach based on questionnaires using recurrent neural network models is proposed for participants’ skin attribute prediction. The prediction engine, which is the most important part of this novel approach, is composed of three prediction models. Each of these models is a neural network allocated to predict different skin attributes: Tone, Spots, and Hydration. We also provide a detailed analysis and solution about the preprocessing of data, the selection of key features, and the evaluation of results. Our prediction system is much faster and more cost effective than traditional clinical skin attribute tests. The system performs very well, and the prediction results show good precision, especially for Tone.","tags":["\"Skin attribute prediction\"","\"Key features\"","\"Neural network\""],"title":"Questionnaires-based skin attribute prediction using Elman neural network","type":"publication"},{"authors":["JiadongYang","Hua Xu","LiPan","PeifaJia","FeiLong","MingJie"],"categories":[],"content":"","date":1293840000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522729,"objectID":"ef9916f0b605d05c5e40c758b51a0b99","permalink":"https://thu-xuhua.github.io/publication/task-scheduling-using-bayesian-optimization-algorithm-for-heterogeneous-computing-environments/","publishdate":"2020-09-19T13:38:48.497342Z","relpermalink":"/publication/task-scheduling-using-bayesian-optimization-algorithm-for-heterogeneous-computing-environments/","section":"publication","summary":"Abstract Efficient task scheduling, as a crucial step to achieve high performance for multiprocessor platforms, remains one of the challenge problems despite of numerous studies. This paper presents a novel scheduling algorithm based on the Bayesian optimization algorithm (BOA) for heterogeneous computing environments. In the proposed algorithm, scheduling is divided into two phases. First, according to the task graph of multiprocessor scheduling problems, Bayesian networks are initialized and learned to capture the dependencies between different tasks. And the promising solutions assigning tasks to different processors are generated by sampling the Bayesian network. Second, the execution sequence of tasks on the same processor is set by the heuristic-based priority used in the list scheduling approach. The proposed algorithm is evaluated and compared with the related approaches by means of the empirical studies on random task graphs and benchmark applications. The experimental results show that the proposed algorithm is able to deliver more efficient schedules. Further experiments indicate that the proposed algorithm maintains almost the same performance with different parameter settings.","tags":["\"Multiprocessor scheduling\"","\"Heterogeneous\"","\"Parallel computing\"","\"Bayesian optimization algorithm\""],"title":"Task scheduling using Bayesian optimization algorithm for heterogeneous computing environments","type":"publication"},{"authors":["Hua Xu"],"categories":null,"content":"Course Classification: Public Elective Courses of Tsinghua University\nLecturer: Hua Xu\nTarget Audience: All Undergraduate Students\nTeaching Time：2011 - Today\n","date":1293811200,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1293811200,"objectID":"3e7631952ecff1e05471581a1af8e97d","permalink":"https://thu-xuhua.github.io/talk/data-mining-method-and-application/","publishdate":"2011-01-01T00:00:00+08:00","relpermalink":"/talk/data-mining-method-and-application/","section":"talk","summary":"Public Elective Courses of Tsinghua University","tags":[],"title":"Data Mining: Methods and Applications","type":"talk"},{"authors":["ZhongwuZhai","BingLiu","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1280620800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523754,"objectID":"83f57721b887d915f9a92e65625776df","permalink":"https://thu-xuhua.github.io/publication/grouping-product-features-using-semi-supervised-learning-with-soft-constraints/","publishdate":"2020-09-19T13:55:53.405011Z","relpermalink":"/publication/grouping-product-features-using-semi-supervised-learning-with-soft-constraints/","section":"publication","summary":"","tags":[],"title":"Grouping Product Features Using Semi-Supervised Learning with Soft-Constraints","type":"publication"},{"authors":["YunWen","Hua Xu","JiadongYang"],"categories":[],"content":"","date":1262304000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523752,"objectID":"9df8a63e1b2dc10a24c56922df802103","permalink":"https://thu-xuhua.github.io/publication/a-heuristic-based-hybrid-genetic-algorithm-for-heterogeneous-multiprocessor-scheduling/","publishdate":"2020-09-19T13:55:51.366191Z","relpermalink":"/publication/a-heuristic-based-hybrid-genetic-algorithm-for-heterogeneous-multiprocessor-scheduling/","section":"publication","summary":"Effective task scheduling, which is essential for achieving high performance of parallel processing, remains challenging despite of extensive studies. In this paper, a heuristic-based hybrid Genetic Algorithm (GA) is proposed for solving the heterogeneous multiprocessor scheduling problem. The proposed algorithm extends traditional GA-based approaches in three aspects. First, it incorporates GA with Variable Neighborhood Search (VNS), a local search metaheuristic, to enhance the balance between global exploration and local exploitation of search space. Second, two novel neighborhood structures, in which problem-specific knowledge concerned with load balancing and communication reduction is utilized, are proposed to improve both the search quality and efficiency of VNS. Third, the use of GA is restricted to map tasks to processors while an upward-ranking heuristic is introduced to determine the task sequence assignment in each processor. Simulation results indicate that our proposed algorithm consistently outperforms several state-of-art scheduling algorithms in terms of the schedule quality while maintaining high performance within a wide range of parameter settings. Further experiments are carried out to validate the effectiveness of the hybridized VNS.","tags":["\"genetic algorithm\"","\"heterogeneous multiprocessor scheduling\"","\"memetic algorithm\"","\"variable neighborhood search\""],"title":"A Heuristic-Based Hybrid Genetic Algorithm for Heterogeneous Multiprocessor Scheduling","type":"publication"},{"authors":["ZhongwuZhai","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1262304000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600522731,"objectID":"080d38026ab842828db02f1173288682","permalink":"https://thu-xuhua.github.io/publication/an-empirical-study-of-unsupervised-sentiment-classification-of-chinese-reviews/","publishdate":"2020-09-19T13:38:50.591349Z","relpermalink":"/publication/an-empirical-study-of-unsupervised-sentiment-classification-of-chinese-reviews/","section":"publication","summary":"This paper is an empirical study of unsupervised sentiment classification of Chinese reviews. The focus is on exploring the ways to improve the performance of the unsupervised sentiment classification based on limited existing sentiment resources in Chinese. On the one hand, all available Chinese sentiment lexicons — individual and combined — are evaluated under our proposed framework. On the other hand, the domain dependent sentiment noise words are identified and removed using unlabeled data, to improve the classification performance. To the best of our knowledge, this is the first such attempt. Experiments have been conducted on three open datasets in two domains, and the results show that the proposed algorithm for sentiment noise words removal can improve the classification performance significantly.","tags":["\"sentiment noise words\"","\"unsupervised sentiment classification\"","\"domain dependent\""],"title":"An Empirical Study of Unsupervised Sentiment Classification of Chinese Reviews","type":"publication"},{"authors":["JiadongYang","Hua Xu","YunpengCai","PeifaJia"],"categories":[],"content":"","date":1262304000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523751,"objectID":"ae598d59064019a767f90eda93e1b207","permalink":"https://thu-xuhua.github.io/publication/effective-structure-learning-for-eda-via-l1-regularizedbayesian-networks/","publishdate":"2020-09-19T13:55:50.396743Z","relpermalink":"/publication/effective-structure-learning-for-eda-via-l1-regularizedbayesian-networks/","section":"publication","summary":"The Bayesian optimization algorithm (BOA) uses Bayesian networks to explore the dependencies between decision variables of an optimization problem in pursuit of both faster speed of convergence and better solution quality. In this paper, a novel method that learns the structure of Bayesian networks for BOA is proposed. The proposed method, called L1BOA, uses L1-regularized regression to find the candidate parents of each variable, which leads to a sparse but nearly optimized network structure. The proposed method improves the efficiency of the structure learning in BOA due to the reduction and automated control of network complexity introduced with L1-regularized learning. Experimental studies on different types of benchmark problems are carried out, which show that L1BOA outperforms the standard BOA when no a-priori knowledge about the problem structure is available, and nearly achieves the best performance of BOA that applies explicit complexity controls.","tags":["\"estimation of distribution algorithms\"","\"bayesian optimization algorithm\"","\"regularization paths\"","\"l1-penalized regression\"","\"bayesian network\""],"title":"Effective Structure Learning for EDA via L1-Regularizedbayesian Networks","type":"publication"},{"authors":["ZhongwuZhai","Hua Xu","JunLi","PeifaJia"],"categories":[],"content":"","date":1262304000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523753,"objectID":"72d1f77f9618e15c8656404e85739c91","permalink":"https://thu-xuhua.github.io/publication/feature-subsumption-for-sentiment-classification-in-multiple-languages/","publishdate":"2020-09-19T13:55:52.371598Z","relpermalink":"/publication/feature-subsumption-for-sentiment-classification-in-multiple-languages/","section":"publication","summary":"An open problem in machine learning-based sentiment classification is how to extract complex features that outperform simple features; figuring out which types of features are most valuable is another. Most of the studies focus primarily on character or word Ngrams features, but substring-group features have never been considered in sentiment classification area before. In this study, the substring-group features are extracted and selected for sentiment classification by means of transductive learning-based algorithm. To demonstrate generality, experiments have been conducted on three open datasets in three different languages: Chinese, English and Spanish. The experimental results show that the proposed algorithm's performance is usually superior to the best performance in related work, and the proposed feature subsumption algorithm for sentiment classification is multilingual. Compared to the inductive learning-based algorithm, the experimental results also illustrate that the transductive learning-based algorithm can significantly improve the performance of sentiment classification. As for term weighting, the experiments show that the ``tfidf-c'' outperforms all other term weighting approaches in the proposed algorithm.","tags":[],"title":"Feature Subsumption for Sentiment Classification in Multiple Languages","type":"publication"},{"authors":["YuanYuan","Hua Xu"],"categories":[],"content":"","date":1251703444,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1251703444,"objectID":"ad0ea045d30af138b1150739b7fa3258","permalink":"https://thu-xuhua.github.io/publication/balancing-convergence-and-diversity-in-decomposition-based-many-objective-optimizers/","publishdate":"2009-08-31T15:24:04+08:00","relpermalink":"/publication/balancing-convergence-and-diversity-in-decomposition-based-many-objective-optimizers/","section":"publication","summary":"The decomposition-based multiobjective evolutionary algorithms (MOEAs) generally make use of aggregation functions to decompose a multiobjective optimization problem into multiple single-objective optimization problems. However, due to the nature of contour lines for the adopted aggregation functions, they usually fail to preserve the diversity in high-dimensional objective space even by using diverse weight vectors. To address this problem, we propose to maintain the desired diversity of solutions in their evolutionary process explicitly by exploiting the perpendicular distance from the solution to the weight vector in the objective space, which achieves better balance between convergence and diversity in many-objective optimization. The idea is implemented to enhance two well-performing decomposition-based algorithms, i.e., MOEA, based on decomposition and ensemble fitness ranking. The two enhanced algorithms are compared to several state-of-the-art algorithms and a series of comparative experiments are conducted on a number of test problems from two well-known test suites. The experimental results show that the two proposed algorithms are generally more effective than their predecessors in balancing convergence and diversity, and they are also very competitive against other existing algorithms for solving many-objective optimization problems.","tags":[],"title":"Balancing Convergence and Diversity in Decomposition-Based Many-Objective Optimizers","type":"publication"},{"authors":["YunpengCai","XiaominSun","Hua Xu","PeifaJia"],"categories":[],"content":"","date":1167609600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1600523755,"objectID":"0286b1471b9d65e2909dba8dd02240b8","permalink":"https://thu-xuhua.github.io/publication/cross-entropy-and-adaptive-variance-scaling-in-continuous-eda/","publishdate":"2020-09-19T13:55:54.412453Z","relpermalink":"/publication/cross-entropy-and-adaptive-variance-scaling-in-continuous-eda/","section":"publication","summary":"This paper deals with the adaptive variance scaling issue incontinuous Estimation of Distribution Algorithms. A phenomenon is discovered that current adaptive variance scaling method in EDA suffers from imprecise structure learning. A new type of adaptation method is proposed to overcome this defect. The method tries to measure the difference between the obtained population and the prediction of the probabilistic model, then calculate the scaling factor by minimizing the cross entropy between these two distributions. This approach calculates the scaling factor immediately rather than adapts it incrementally. Experiments show that this approach extended the class of problems that can be solved, and improve the search efficiency in some cases. Moreover, the proposed approach features in that each decomposed subspace can be assigned an individual scaling factor, which helps to solve problems with special dimension property.","tags":["\"cross entropy\"","\"adaptive variance scaling\"","\"estimation of distirbution algorithms\"","\"evolutionary computation\""],"title":"Cross Entropy and Adaptive Variance Scaling in Continuous EDA","type":"publication"},{"authors":["JunhuiDeng"],"categories":null,"content":"Course Classification: Tsinghua University Computer Department Undergraduate Professional Basic Course\nLecturer：Junhui Deng\nTarget Audience: Computer Science Undergraduate\nTeaching Time：2001 - 2006\n","date":978278400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":978278400,"objectID":"b7caabb0e4bde45ecc0519114b6a051c","permalink":"https://thu-xuhua.github.io/talk/data-structure-cs/","publishdate":"2001-01-01T00:00:00+08:00","relpermalink":"/talk/data-structure-cs/","section":"talk","summary":"Tsinghua University Computer Department Undergraduate Professional Basic Course","tags":[],"title":"Data Strcture","type":"talk"},{"authors":["JunhuiDeng"],"categories":null,"content":"Course Classification: Public Elective Courses of Tsinghua University\nLecturer：Junhui Deng\nTarget Audience: All Undergraduate Students\nTeaching Time：2002 - Today\n","date":978278400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":978278400,"objectID":"970c56d891d3e1abdca0aaf1b5140c8d","permalink":"https://thu-xuhua.github.io/talk/data-structure/","publishdate":"2001-01-01T00:00:00+08:00","relpermalink":"/talk/data-structure/","section":"talk","summary":"National Excellent Course, Public Elective Course of Tsinghua University","tags":[],"title":"Data Strcture","type":"talk"},{"authors":["JunhuiDeng","Hua Xu"],"categories":null,"content":"Course Classification: Tsinghua University Computer Department Graduate Basic Theory Course\nLecturer：Junhui Deng, Hua Xu\nTarget Audience: All Graduate Students\nTeaching Time：1997 - Today\n","date":852048000,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":852048000,"objectID":"f93308c8be1670e9b2dc795f64282a4a","permalink":"https://thu-xuhua.github.io/talk/computational-geometry/","publishdate":"1997-01-01T00:00:00+08:00","relpermalink":"/talk/computational-geometry/","section":"talk","summary":"Tsinghua University Computer Department Graduate Basic Theory Course","tags":[],"title":"Computational Geometry","type":"talk"},{"authors":["YuanYuan","Hua Xu"],"categories":[],"content":"","date":-62135596800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":-62135596800,"objectID":"6b7b52354494c1b60185acf84b1a8696","permalink":"https://thu-xuhua.github.io/publication/a-new-dominance-relation-based-evolutionary-algorithm-for-many-objective-optimization/","publishdate":"0001-01-01T00:00:00Z","relpermalink":"/publication/a-new-dominance-relation-based-evolutionary-algorithm-for-many-objective-optimization/","section":"publication","summary":"Many-objective optimization has posed a great challenge to the classical Pareto dominance-based multiobjective evolutionary algorithms (MOEAs). In this paper, an evolutionary algorithm based on a new dominance relation is proposed for many-objective optimization. The proposed evolutionary algorithm aims to enhance the convergence of the recently suggested nondominated sorting genetic algorithm III by exploiting the fitness evaluation scheme in the MOEA based on decomposition, but still inherit the strength of the former in diversity maintenance. In the proposed algorithm, the nondominated sorting scheme based on the introduced new dominance relation is employed to rank solutions in the environmental selection phase, ensuring both convergence and diversity. The proposed algorithm is evaluated on a number of well-known benchmark problems having 3-15 objectives and compared against eight state-of-the-art algorithms. The extensive experimental results show that the proposed algorithm can work well on almost all the test functions considered in this paper, and it is compared favorably with the other many-objective optimizers. Additionally, a parametric study is provided to investigate the influence of a key parameter in the proposed algorithm.","tags":[],"title":"A New Dominance Relation Based Evolutionary Algorithm for Many-Objective Optimization","type":"publication"}]