The increasing pace of digitalization and globalization has led to a significant increase in the amount of information available in the field of science, including psychology. To help researchers and other interested parties navigate this information overload, we developed PsychTopics, an online tool that uses machine learning to identify research topics and trends in psychology literature from the German-speaking countries (www.psyndex.de/en). The topics are continuously detected with RollingLDA – a variant of topic modeling that enables sequential modeling of dynamically growing corpora and ensures time consistency of resulting time series.
📜 Find the research paper here:
Bittermann, A. & Rieger, J. (2022). Finding Scientific Topics in Continuously Growing Text Corpora. In Proceedings of the Third Workshop on Scholarly Document Processing, pages 7–18, Gyeongju, Republic of Korea. Association for Computational Linguistics. https://aclanthology.org/2022.sdp-1.2
https://abitter.shinyapps.io/psychtopics/
- Start R
- Run
shiny::runGitHub('psychtopics','leibniz-psychology') - Important when using RStudio: In the new RStudio window, click "Open in Browser" in the top menu
- The app should be up and running in your browser!
You may also need to the install following packages:
install.packages(c('devtools','config','echarts4r','golem','htmlwidgets','reactable', 'shiny', 'shiny.react','shiny.router','tsbox', 'spsComps'))
devtools::install_github('Appsilon/shiny.fluent')