基于语义理解的中文博文倾向性分析

Orientation analysis for Chinese blog text based on semantic comprehension

博客作为一种大众化的信息及文化载体被越来越多的人所接受,博客文本的情感倾向性分析也逐渐成为信息挖掘领域的热点。目前,文本倾向性分析的研究大都围绕普通文本、新闻评论进行,针对博客文本的特点,提出一种基于语义理解的博客文本倾向性分类方法。首先以HowNet情感词语集为基准,构建中文基础情感词典,并用中文词语相似度方法计算词语的情感权值,同时分析语义层副词的出现规律及其对文本倾向性判断的影响,最后利用博主的语言风格因素对倾向性结果进行修正实现博文的情感分类。实验表明,该方法能有效地判定博客文本情感倾向性。

Blog has been accepted by more and more people as a popular information and cultural carrier.Orientation analysis for blog text also has become a hot spot in the field of information mining.The previous researches of text orientation mainly focus on plain text or news comments.A method of orientation analysis for blog text based on semantic comprehension was proposed according to the characteristics of blog text.Firstly,a Chinese basic emotional lexicon dictionary based on the HowNet emotional word set was constructed and the emotional value of Chinese emotional words was calculated on the basis of the similarity of Chinese words.Then,the adverbs and its influence on identification of text orientation in the semantic level were analyzed.Finally,the results were amended by using bloggers＇ language style factors and then the sentimental classification for blog text was realized.The experimental results show that the proposed method can effectively judge the blog text sentimental preference.