Text clustering method based on topic model

The invention discloses a text clustering method based on a topic model. The text clustering method comprises the following steps: acquiring a to-be-tested document; presetting a document topic model, carrying out parameter solving on the document topic model, and constructing a dynamic evolution model of the emotion polarity characteristics along with time; and mining the evolution process of words and emotions under the corresponding themes. The invention provides a text clustering method based on a topic model, which comprises the following steps of: mining polarity characteristics of topics and emotions, capturing a dynamic evolution relationship between the topics and the emotions, and modeling in combination with a time sequence to reflect emotional evolution situations of people aiming at different topics under the time sequence; and the emotion trend of the public for a certain theme in a certain period can be better analyzed. 本发明公开了一种基于主题模型的文本聚类方法，包括：获取待测的文档；预置文档主题模型，并对文档主题模型进行参数求解，构建情感极性特征随时间的动态演化模型