融合情感词典和语义规则的微博评论细粒度情感分析

Fine-grained Sentiment Analysis of Microblog Comments Based on Fusion of Sentiment Lexicon and Semantic Rules

[目的/意义]旨在为政府精准引导舆情提供参考。[方法/过程]在情感词汇本体库的基础上,扩充表情符号、网络用语、单字情感词等多类情感词典,结合语义规则建立了微博情感分析模型。同时,通过对比词频模型验证了模型的有效性,以重庆万州公交车坠江事件为例验证了模型的实用性。[结果/结论]提出的情感分析模型较词频模型有效提升了细粒度情感分类的准确率,使用该方法分析舆情期间情感的演化能够帮助政府实时掌握舆情动态,有助于政府部门实现舆情的应急管理和有效控制。

[Purpose/significance]The paper is to provide references for the government to accurately guide public opinion.[Method/process]On the basis of the sentiment words ontology base the paper expands the sentiment lexicons of emotion icons network terms single-word sentiment words establishes the emotional analysis model of microblog by semantic rules verifies the validity of the model by comparing the word frequency model and verifies the practicability of the model by taking the event of“A bus falls into a river in Wanzhou Chongqing”as an example.[Result/conclusion]Compared with the word frequency model the proposed emotional analysis model can effectively improve the accuracy of fine-grained sentiment classification.Using this method to analyze the emotional evolution during the public opinion period can help the government to grasp the public opinion dynamics in real time and it is helpful for government departments to achieve emergency management and effective control of public opinion.