论文部分内容阅读
一、研究背景近年来BBS的迅速发展,主观性的言论越来越多。如何构建一个高效的系统对如此丰富的信息资源进行分析和处理,成为一个重要的研究问题。而对BBS信息资源的分析和处理,可以通过中心词和情感词两个维度进行处理。对于中心词的归类已经有了相对较为完整的分析体系,产生了很多有监督的学习方法以及文本特征表示方法和特征选择机制。而情感词的分类和评级以主观词为主,因此针对中心词的选择机制及方法在情感词上不能完全加以应用。本文主要以中山大学BBS的帖子为研究内容,选择合适的情感词分类方法,对情感词进行细分及评级,从而得出一个满足中大BBS舆情分析系统的情感词分类体系。
First, the research background In recent years the rapid development of BBS, subjective speech more and more. How to build an efficient system to analyze and deal with such a rich information resources has become an important research issue. However, the analysis and processing of BBS information resources can be handled through the two dimensions of center words and emotion words. There has been a comparatively complete analysis system for the categorization of central words, resulting in a lot of supervised learning methods as well as text feature representation and feature selection mechanisms. The classification and rating of emotional words are mainly subjective. Therefore, the selection mechanism and method of central words can not be completely applied to the emotional words. This article mainly takes the post of BBS of Sun Yat-sen University as the research content, selects the appropriate method of emotion word classification, subdivides and grades the emotion word, and gets a emotion word classification system that satisfies the BBS public opinion analysis system of Zhongda University.