基于电影网站短评数据的网络舆情文本挖掘与情感分析  

Text Mining and Emotion Analysis of Online Public Opinion Based on Short Review Data of Film Websites

作　　者：贺海玉[1] HE Haiyu(Information Technology Department of Dazhong Newspaper Group,Ji'nan 250014,China)

机构地区：[1]大众报业集团信息技术部,山东济南250014

出　　处：《现代信息科技》2023年第21期126-130,135,共6页Modern Information Technology

摘　　要：对电影短评数据进行情感分析的目的是为了获取观众对某部电影的情感倾向,同时还可帮助电影制作者通过了解观众的情感倾向,从而改善电影的制作。文章采用的方法是通过Python代码爬取电影网站上的评论数据,对爬取的数据进行多项数据预处理技术得到较为规范的评论数据,再利用TF-IDF算法计算出短评数据的关键词及权重并给关键字词云图,然后使用SnowNLP库计算出短评数据的情感分值,并运用LDA模型对电影网站短评数据主题分类,最终给出电影网站短评数据情感分析的可视化评价结果。The purpose of conducting emotion analysis on film short review data is to obtain the audience's emotional tendencies towards a certain film.At the same time,it can help filmmakers improve film production by understanding the audience's emotional tendencies.The method used in this paper is to crawl the review data on film websites through Python code,perform multiple data preprocessing techniques on the crawled data to obtain more standardized review data,then use TF-IDF algorithm to calculate the keywords and weights of the short review data and give keywords cloud maps.Then,it uses SnowNLP library to calculate the emotional score of the short review data,and uses LDA model to classify the short review data theme of the film websites.Finally,it provides a visual evaluation result of emotion analysis of short review data on film websites.

关 键 词：情感分析 Jieba分词 TF-IDF算法 SnowNLP分析 LDA主题模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]