An Efficient Multi-Dimensional Sentiment Analysis Method Based on Text Mining

Traditional manual methods can no longer meet the needs of analyzing the increasing number of literary works, and most existing sentiment analysis technologies are limited to simple sentiment classification and cannot capture the rich emotional expression levels and complex connotations in literary works. To address this problem, this paper proposes a multi-dimensional sentiment analysis method based on text mining. This method first preprocesses the input text to reduce noise and highlight key information. Then, sentiment latent Dirichlet allocation (SLDA) topic clustering and sentiment classification are performed to automatically discover potential topics and sentiment tendencies in the text through the probability distribution of words. In order to enhance data representation, this paper also uses Word Cloud generation technology and keyword co-occurrence graph. Finally, the multi-dimensional processing results are delivered to AI Agent for sentiment analysis. Experimental results show that compared with traditional sentiment classification methods such as LDA, this method can efficiently and accurately obtain sentiment analysis descriptions of large-scale texts, showing obvious superiority. This study also developed a graphical user interface (GUI) that integrates the above complete method to achieve end-to-end multi-dimensional automated sentiment analysis.