Research on semantic sentiment analysis of Chinese short-text

被引：0

作者：

Yu, Jian ^{[1
]}

Gao, Jie ^{[1
]}

Yu, Mei ^{[1
]}

Han, Xu ^{[1
]}

Zhang, Xu ^{[1
]}

机构：

[1] Tianjin Key Laboratory of Cognitive Computing and Application, School of Computer Science and Technology, Tianjin University, No. 92, Weijin Road, Nankai District, Tianjin, China

来源：

ICIC Express Letters | 2015年 / 9卷 / 12期

关键词：

Semantics - Data mining;

D O I：

暂无

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

The environment in social network is diversified and open. Due to the semantic ambiguity of Chinese over English, the sentiment analysis methods of English are inappropriate for Chinese text. Thus, we need to analyze the sentiment of Chinese text according to the feature of Chinese short-text. In this paper, we first propose PPSA, Phrase Pattern Sentiment Analysis, which fully uses the semantic lexicon and extends the traditional rules of phrase pattern. Furthermore, WSSA is presented, Word Statistics Sentiment Analysis, which combines the analysis of sentence structure based on PPSA. The experiment results reveal that the accuracy of WSSA is above 6% higher than PPSA. © 2015 ICIC International.

引用

页码：3237 / 3244