Тематическое моделирование текстовых учебных материалов по информатике средствами языка R

This paper presents results of topic modeling for text learning materials. Learning materials are electronic lecture notes used by teachers to prepare for computer science classes. Topic modeling methods allow users to systematize the content of textual documents without additional manual work. Main topics in documents are highlighted, and the distribution of topics in documents is demonstrated. In other words, the proposed methods provide the framework for the so-called topic model that puts a set of topics that characterize the content of documents in a given collection of documents. The latent Dirichlet allocation (LDA) is used for topic modeling. The implementation is done using the R language. The developed interactive web application provides a set of visual tools for topic modeling to a user (teacher). Visualization techniques gradually improve the ergonomics of a teacher’s work with learning materials and save the time spent on studying, analyzing, and selecting relevant study materials.DOI 10.14258/izvasu(2018)4-12