计算机应用与软件Issue(12)：240-243,4.DOI:10.3969/j.issn.1000-386x.2013.12.063
基于短文本及情感分析的微博舆情分析
ANALYSING MICROBLOGGING PUBLIC OPINIONS BASED ON SHORT TEXT AND SENTIMENT ANALYSIS
李岩 1韩斌 1赵剑1
作者信息
- 1. 江苏科技大学计算机科学与工程学院 江苏 镇江 212000
- 折叠
摘要
Abstract
Aiming at the characteristics represented in microblogging text such as incomplete property , sparseness and fragmentation , we design a microblogging public opinion system which is based on short text clustering and the sentiment analyses of user comments .Through extracting the session of phrases and messages ’ flow, the system solves the problems of incomplete and overlapping properties in short messages.Through improving the TF-IDF, the system solves the problem of similarity drift caused by sparseness of keywords .The hybrid short text clustering algorithm designed can effectively cluster the microblogging text messages and can meet the requirements of accuracy and scalability.The sentiment words analyses method can directly analyse against the comments on hot topics , identify their emotional tendencies and the evolution rule , and understand the viewpoints and attitudes of internet users on hot topics further .In end of the paper we verify the feasibility of the method and the experimental results through experiments .关键词
微博/热点事件/短文本聚类/情感分析Key words
Microblog/Hot events/Short text clustering/Sentiment analysis分类
信息技术与安全科学引用本文复制引用
李岩,韩斌,赵剑..基于短文本及情感分析的微博舆情分析[J].计算机应用与软件,2013,(12):240-243,4.
