Emotion Recognition of College Students’ Online Learning Engagement Based on Deep Learning

Abstract

In actual learning scenarios, learners have more and more personalized needs. The traditional measuring tools of emotional engagement can no longer meet the personalized needs of online learning. To solve the problem, this paper explores the emotion recognition of college students' online learning engagement based on deep learning. Firstly, the features were extracted from the texts related to online learning reviews and interactive behaviors of college students, and the texts were vectorized by the multi-head attention mechanism. Based on the multi-head attention mechanism, a bidirectional long short-term memory (BLSTM) emotion classification model was established, which describes the emotional attitude of learners towards learning engagement more clearly and more accurately. Through experiments, the proposed model was proved effective in emotion recognition of college students’ online learning engagement.