Text Sentiment and Semantic Networks Analysis for MOOC Review Texts

This paper uses web crawler technology to retrieve the review data of six high-quality courses in MOOC English. Review texts are pre-processed and sentiment analysis is performed by using ROSTCM6. After that, TF-IDF is used to calculate the vector matrix, and the number of clusters is determined by the “elbow method”. Finally, K-means is used to cluster negative sentiment comments. Each category is visualized and analyzed for feature extraction. Based on sentiment analysis, three suggestions for MOOC English courses are finally put forward.