Prediction API

Introduction

AMiner Prediction API is a toolkit for science data prediction, such as scholar portrait property prediction. The toolkit aims to utilize science data and machine learning algorithms to provide more intelligent functionality for global researchers. All algorithms and models which the toolkit uses are derived from AMiner.

Pre-requirement

Anaconda is strongly recommended for environment configuration. Additionally, some libraries are requied. Use following commands to install these libraries:

Library	Command
fastText	conda install -c mbednarski fasttext *****
Scikit-learn	conda install -c anaconda scikit-learn
Jieba	conda install -c conda-forge jieba
Requests	conda install -c conda-forge requests
Tensorflow	conda install -c conda-forge tensorflow
Pytorch	conda install -c pytorch pytorch
Numpy	conda install -c conda-forge numpy
Pandas	conda install -c conda-forge pandas
BeautifulSoup4	conda install -c conda-forge beautifulsoup4
Scrapy	conda install -c conda-forge scrapy
Levenshtein	conda install -c conda-forge python-levenshtein

***** If you are using OSX, you should use following command to intall fastText

conda install -c conda-forge fasttext

Models Download

Toolkit depends on some pre-trained model files which can be downloaded at following address:

Download

Extract it and move all files into model directory before testing code.

Document

NSFC Subject Classifier

NSFC AI Subject Classifier

Gender Prediction

Identity Prediction

Jobhopping Prediction

Expert Recommendation

Paper Ranker