gativetendencies,soastocarryoutthecorrespondingtextmining.
It feels a little abstract, and my personal interpretation of it is that
when we take a piece of text or many pieces of text, we do a
sentiment analysis of the text, and we say, "This product is great,"
then we think that the text is positive. It's simple. However,
with the continuous progress of science and technology, there are
more and more comments, and we cannot read them completely. At
this time, we can extract a large number of text topics, which can be
used for information mining. At the same time, if we want to know
both positive and negative themes in a large text, then we can do
sentiment analysis. Conduct sentiment analysis to determine user
reviews of products. To deal with and analyze the problems of good
目 录
摘要.....................................................................................................................................................1
一 绪论................................................................................................................................................3
1.1 研究目的和意义..........................................................................................................3
二 国内外研究现状............................................................................................................................4
三 研究内容........................................................................................................................................6
3.1 定义挖掘目标:...........................................................................................................6
3.2 定义挖掘分析步骤:...................................................................................................6
四 技术路线(工作流程)................................................................................................................7
4.1 爬取数据......................................................................................................................7
4.2 数据清洗.......................................................................................................................8
五、构建模型...................................................................................................................................18
5.1 评论数据情感倾向分析.............................................................................................18
5.2 修正情感倾向.............................................................................................................23
5.3 查看情感分析效果.....................................................................................................28
5.4.使用 LDA 主题模型进行主题分析...........................................................................34
六 总结......................................................................................................................................43
七 附录..............................................................................................................................................45
大雅相似度查询(学习通提供技术支持)