计算机应用与软件2016，Vol.33Issue(10)：167-171,5.DOI:10.3969/j.issn.1000-386x.2016.10.037
汉语组块分析在情感分类中的应用研究
ON APPLYING CHINESE CHUNK PARSING IN SENTIMENT CLASSIFICATION
摘要
Abstract
The sentiment analysis of online product reviews plays an important role in decision-making of Internet users’daily purchase behaviour,therefore,the way to well use fine-grained processing method in improving the accuracy of sentiment analysis becomes a hot research topic.Aiming at this issue,the paper proposes a Chinese chunk parsing-based emotion recognition method.First,it relies on Chinese chunk parsing to make fine-grained processing on car reviews corpus and extracts the emotion labels as well.Then,it combines sentiment words ontology and support vector machine model to classify emotion labels so as to implement the discrimination of emotional orientation.It is demonstrated by experiment that compared with other classification algorithms,the sentiment classification method using Chinese chunk parsing improves the average accuracy by 4%.Therefore the sentiment classification based on Chinese chunk parsing can reduce the input feature dimensions and effectively improve the performance of classifier.关键词
汉语组块分析/情感标签/情感词本体/情感分类Key words
Chinese chunk parsing/Emotion label/Sentiment words ontology/Sentiment classification分类
信息技术与安全科学引用本文复制引用
杜思奇,李红莲,吕学强..汉语组块分析在情感分类中的应用研究[J].计算机应用与软件,2016,33(10):167-171,5.基金项目
国家自然科学基金项目（61271304）；北京市教委科技发展计划重点项目暨北京市自然科学基金 B 类重点项目（KZ201311232037）。 （）
