Sentiment Analysis of Animated Online Education Texts Using Long Short-Term Memory Networks in the Context of the Internet of Things

Abstract

This work aims to introduce Long Short-Term Memory (LSTM) under the Internet of Things (IoT) context to enhance the accuracy and granularity of sentiment analysis in animated online education texts. It employs a multimodal data collection approach and uses IoT technology to gather multimodal textual data from students engaged in animated online education. The data includes students' feedback texts, emotional texts, written texts, and verbal expressions during animated online education. Subsequently, a model named Information Block Bidirectional Long-Short term Memory (IB-BiLSTM) is designed and utilized to construct a sentiment classification model for animated online education texts. Experimental results demonstrate that the model achieves an accuracy of 93.92% and an F1-score of 90.34% for sentiment classification in animated online education texts and the loss function converges to around 0.14. This model effectively captures the emotional changes and evolution during students' learning process. Thus, the proposed model holds significant potential and practical significance for enhancing animated online education's personalization and emotional engagement. It provides valuable insights and guidance for the intelligent development of the education field.