Identifying WEB Public Sentiment of Opinions

Abstract—One can express personal experiences and opinions on almost anything, at review sites, forums, discussion groups, blogs, and etc. As the Internet is growing rapidly, web public sentiment can reflect the majority’s attitude to some opinions, which pertain to a specific dimension (topic) of the subject domain. Quotes of the attitude may be generally positive or negative. In this paper we present a new approach that employs TF-PDF and clustering analysis for opinions extraction and phrase patterns for the calculation of their sentiment orientation of opinions using the multilayer neural network. The experimental result shows that the approach is qualitatively quite useful when used to analyze the netizens ’ opinions to Sanlu tainted milk event from some websites. Index Terms—opinion mining;sentiment analysis, TF/PDF, BP neural network I