Sentiment Analysis of Micro-blog Integrated on Explicit Semantic Analysis Method

TLDR: After the introduction of Wikipedia to micro-blog text feature expansion, the final evaluation index of the classification results of the naive Bias classifier has been improved, which achieves a better classification effect and improves the effectiveness of the sentiment classification.

Abstract: Combined with existing research of short text classification, this paper analyzes and explores the structure and characteristics of Wikipedia. And the fusion of explicit semantic analysis algorithm micro-blog sentiment analysis method is proposed. Wikipedia is regarded as external semantic knowledge base, and the entries are introduced as a supplement of micro-blog text features. The author improves the previous micro-blog sentiment analysis text representation method, and then constructs the naive Bias classifier to achieve emotion classification. The experimental results show that after the introduction of Wikipedia to micro-blog text feature expansion, the final evaluation index of the classification results of the naive Bias classifier has been improved, which achieves a better classification effect and improves the effectiveness of the sentiment classification.