摘要微博情绪分析旨在对微博文本表达的情绪信息进行自动分析，这是自然语言处理研 究领域的一个热点问题。对于给定的一条微博文本,它将输出该条微博的情绪类别,比如, 喜好、惊讶、愤怒等感觉。74312

本文为研究适用于微博文本的情绪分类体系，进行了情绪词典构建，构造了表情词 典、卡方词典，实现了基于规则的情绪分类方法。采用朴素贝叶斯、SVM、KNN、MLKNN 算法，实现了基于机器学习的情绪分类方法。进行序列模式挖掘，实现了结合规则和机 器学习的情绪分类方法。并且以 NLP&CC2013 提供的情绪语料为实验对象，验证了实现 的情绪分类方法的有效性。

毕业论文关键词 情绪分类 情绪词典 机器学习 序列模式挖掘

毕业设计说明书外文摘要 

Title Topic-specific sentiment analysis  on microblogging

Abstract Microblog sentiment analysis is designed to automatically analyze the emotional information expressed by microblog, which is a hot issue in the field of Natural Language Processing research。 Given a microblog text, we classify its emotion into many emotional categories, for example, like, surprise, anger and so on。 In order to study the sentiment classification system applied to microblog, this paper constructs emotional dictionaries。 For instance, a facial expression dictionary and a chi square dictionary。 Then this paper implements the method of emotion classification based on rules。 In addition, the method of emotion classification based on machine learning is realized by using the Naive Bayes, SVM, KNN and MLKNN algorithms。 Sequential pattern mining is carried out, and the emotion classification method of combining rules and machine learning is realized。 This paper uses the emotional data provided by NLP&CC2013 as the experimental object, which verifies the effectiveness of the emotional classification methods mentioned in the paper。

Keywords Emotion Classification Emotion Dictionary Machine Learning Sequential pattern mining

目 次

1 绪论 1

1。1 研究背景 1

1。2 研究意义 1

1。3 情绪分类研究现状 2

1。4 主要研究内容和组织结构 2

2 基于规则的情绪分类 4

2。1 情绪词典 DUTIR 4

2。2 微博情绪词典构建 4

2。3 基于规则的情绪分类方法 6

2。4 实验数据以及评价指标 6

2。5 实验结果以及实验分析 7