二元情感常识库建设及其在文本情感分析中的应用
首发时间:2013-01-05
摘要:常识知识一直是一个研究热点,情感常识作为情感计算领域的一个重要方面,对文本情感分析有重要作用。本文对现有中文情感常识库进行结构优化和扩展,以二元结构表示一条常识内容,这些二元结构包括动宾结构、主谓结构、情感搭配以及产品属性和评价词组合等多种形式,通过人工标注其情感类别,建立了一个新型二元情感常识库。借助该情感常识库在情感语料库和公共语料集上进行文本情感分析实验和倾向性分析实验,实验证明该情感常识库可以有效辅助进行文本情感分析和倾向性分析。
关键词: 自然语言处理 情感分析 情感常识 二元情感常识库 倾向性分析
For information in English, please click here
Construction of the Binary Affective Commonsense Knowledgebase and its Application in Text Affective Analysis
Abstract:Commonsense knowledge has always been a hot research topic. Affective commonsense knowledge, as an important part in affective computing, plays an important role in affective analysis of the text. This paper optimized the structure of existing affective commonsense knowledgebase and expanded its scale, constructed a novel affective commonsense knowledgebase. Each common sense was expressed as a dual structure, which contains VO construction, SV construction, affective collocation, combination of product feature and opinion word, which are finally tagged with affective classification artificially. At last, affective analysis experiment and opinion analysis experiment have been done on emotion corpus and public corpus, and the experiment results show that the affective commonsense knowledgebase is helpful to affective analysis and opinion analysis.
Keywords: Natural Language Processing Affective Analysis Affective Commonsense Knowledge Binary Affective Commonsense Knowledgebase Opinion Analysis
基金:
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
二元情感常识库建设及其在文本情感分析中的应用
评论
全部评论0/1000