基于社会化媒体的情感倾向分析系统的设计与实现

发布时间：2018-11-10 11:00

【摘要】：随着Web2.0以及移动互联网的飞速发展,各种新型应用催生了大量的用户评论数据,而这些评论数据大多包含了用户的观点或情感倾向。对用户的评论数据进行情感倾向性分析,对企业调研用户对产品和服务的反馈,以及政府了解人们对热点事件的看法和意见有着重要的参考作用。情感倾向分析的目的就是挖掘出用户在文本中表达出的对某个实体的情感倾向。本文首先介绍了情感倾向分析的相关研究内容以及研究现状。然后对文本情感倾向分析的算法原理进行了阐述。该文本情感倾向分析算法原理包括两个层次：第一层次是情感词级别的情感极性分类算法,本文通过设计并实现基于表情符号的情感词极性分类算法,利用表情符号的情感倾向和情感词与表情符号之间的关联关系,将情感词的情感倾向分类为正向积极和负向消极两类；第二层次是文本级别的情感倾向分析算法,通过对文本建立基于情感词的情感倾向模型,并对文本的主客观分类以及对主观性文本的情感倾向进行分类,判别出文本的情感倾向是正向还是负向的。在完成了情感倾向分析的核心算法之后,本文基于该核心算法设计并实现了一个基于社会化媒体的情感倾向分析系统(Sentiment Analysis System),简称SA系统。该SA系统从人们针对微博话题所发表的大量微博中分析并统计该微博话题总的情感倾向,并且将情感倾向分析的结果通过Web页面的可视化方式进行了展示。本文从SA系统需求分析、架构设计、数据库设计以及前端页面设计等方面进行阐述,并对实现后的SA系统进行了测试。
[Abstract]:With the rapid development of Web2.0 and mobile Internet, a lot of user comments are generated by various new applications, and most of these comments data contain users' views or emotional tendencies. It is important to analyze the emotional tendency of the users' comment data, to investigate the users' feedback to the products and services, and to understand people's views and opinions on the hot issues. The purpose of affective tendency analysis is to find out the emotional tendency expressed by the user in the text. This paper first introduces the related research content and research status of affective tendency analysis. Then the text emotional tendency analysis algorithm principle is expounded. The principle of the text affective tendency analysis algorithm includes two levels: the first level is the affective polarity classification algorithm at the affective word level. This paper designs and implements the affective word polarity classification algorithm based on emoji. By using the affective tendency of emoji and the relation between emotive words and emoji, the affective tendency of emoticons is classified into positive and negative. The second level is the text level of emotional tendency analysis algorithm, through the text based on affective word emotional preference model, and the text of subjective and objective classification and subjective text emotional tendency to classify. Judge whether the emotional tendency of the text is positive or negative. After completing the core algorithm of affective orientation analysis, this paper designs and implements a social-media-based emotional orientation analysis system, (Sentiment Analysis System), for short, SA system based on the core algorithm. The SA system analyzes and statistics the general emotional tendency of Weibo topic from a large number of topics published by people, and displays the results of emotional tendency analysis through the visualization of the Web page. This paper describes the requirements analysis, architecture design, database design and front-end page design of SA system, and tests the SA system after implementation.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.1;TP393.09

【相似文献】