Categorization and Monitoring of Internet Public Opinion Based on Latent Semantic Analysis

Rapid progress of network arouses much attention on Internet public opinion. To address this issue, we propose a novel system for categorization and monitoring of Internet public opinion. Due to the text format of Internet public opinion and the semantic relationship between words in such documents, we introduce latent semantic analysis (LSA) to represent document of public opinion. Compared to the traditional vector space model (VSM), LSA overcomes the problem of high dimensional space. We use two classifiers to perform text categorization on a corpus collected from a hot Website. For the monitoring, we give the structure of this module and introduce its main functions.