摘要近年来网络时代的繁荣,极大地丰富了人们的生活,越来越多的用户乐于在网络中留下自己的评论和观点。通过对这些评论进行情感分析,可以进一步的优化企业自身的产品,对客户关系管理和舆情控制有很大帮助,目前对于中文的情感分析还处于初级阶段,大多数情感分析成果都是针对英文的,我们可以对外国学者研究成果归纳总结,提出更合适的算法。情感分析是判别文本的情感倾向,主要可以分为正面(褒)、负面(贬)、中性。本文对文本进行分词等预处理,对情感特征采用卡方统计进行计算提取,测试了多个机器学习算法在情感分析中的表现,对目前开发的软件所存在的问题进行归纳总结。82919
毕业论文关键词 情感分析;机器学习;文本分类
毕业设计说明书(论文)外文摘要
Title Research and Application of Dialogue-Based Sentiment Analysis
Abstract In recent years, the prosperity of Internet had greatly enriched people’s life。 More and more netizens are willing to pose their comments or views in various websites。 Sentiment analysis is the technique used to determine the emotional tendency of text。 It generally can be pided by three aspects: positive and negative and neutral。 Through analyzing the sentiment of these comments, the products of the enterprise can be optimized, which is of great help to control the customer relationship management and public opinion。 At present, the Chinese sentiment analysis is still in its infancy, most results of sentiment analysis algorithms aim at English texts。 By summing up the research results of foreign scholars, we can propose an appropriate analysis algorithms suitable for Chinese texts。
In this paper, we carry out the pre-processing, including word segmentation and stop word using, etc。, and the emotional feature extraction with Chi-square statistic method。 We also test the expression of several machine learning algorithms for sentiment analysis, and finally summarize the existing problems of current popular software developments。
Keywords sentiment analysis; machine learning; text classification
目次
1 绪论 1
1。1 研究背景 1
1。2 研究目的 1
1。3 文章结构 2
2 相关介绍与理论概述 2
2。1 情感分析理论概述 2
2。2 研究现状 3
3 软件系统设计 5
3。1 软件设计流程 5
3。2 软件开发技术 6
4 文本预处理技术 6
4。1 中文分词 6
4。2 去除停用词 7
4。3 情感信息抽取 7
4。4 训练语料标记 8
5 情感分类功能设计与实现 9
5。1 情感分类功能设计 9
5。2 朴素贝叶斯 9
5。3 支持向量机 11
5。4 K-最临近算法 14
6 算法测试