Abstract
This paper uses Python technology to obtain relevant data from the Douban platform, analyze user data, urban distribution
and other information, and uses LDA model to build models, train models, analyze and evaluate the emotional trends in user film reviews, including theme confusion and theme consistency. The analysis results help to better understand the habits of audiences in different regions, subjective emotions about film and television works, and have certain reference significance for evaluating and adjusting film market projects.
|