Opinion Mining for Modeling User Experience of Online Education: Sentiment Analysis and Keywords Extraction of Student Reviews

Abstract

The paper discusses the possibilities of applying modern natural language processing technologies of opinion mining to investigate and improve the user experience of online-courses students. We analyzed 27 000 student reviews of projects within the Python programming language course. First, we applied keyword extraction algorithms as a way of semantic compression to receive a generalized picture of what users main impressions are. Then we performed sentiment analysis to understand the feelings of students towards the learning process. The used methodology proved to be effective for analyzing user experience and allowed to find out some discrepancies between information in project descriptions and what users' reflection on the project. Two instruments of SA were applied to receive data on users' feelings in general.

Keywords

• natural language processing
• computational linguistics
• sentiment analysis
• keywords extraction
• user experience
• online courses
• online education