引入营销效应的文本情感挖掘研究
发布时间:2018-02-09 06:07
本文关键词: 在线评论 文本挖掘 情感词典 情感效应 营销效应 出处:《统计与决策》2017年09期 论文类型:期刊论文
【摘要】:大量的在线评论文本数据,对汽车制造业的产品创新、精准营销具有重要的实用价值。传统的文本挖掘多关注情感极性强度而忽视了情感营销效应,挖掘结果与企业需求差异较大。文章利用大连理工、中国知网等情感词典和丰富的汽车语料库,采取了基于CRF机器学习的情感词自动获取方法,构建了情感营销效应词库,并运用MATLAB对该词库情感营销效应进行了全面的统计分析,分类优化情感营销效应值,最后采用中科院ICTCLAS中文分词技术进行测试,结果表明,考虑情感营销效应的文本挖掘召回率达到80%以上,能较好地满足汽车厂商的应用需求。
[Abstract]:A large amount of online review text data is of great practical value for automotive manufacturing product innovation and precision marketing. Traditional text mining pays more attention to emotional polarity and neglects emotional marketing effect. The result of mining is quite different from the demand of enterprise. This paper adopts the automatic acquisition method of emotion words based on CRF machine learning and constructs the affective marketing effect lexicon by using the emotion dictionary such as Dalian Polytechnic Technology, China Zhiwang and the abundant automobile corpus, which is based on CRF machine learning. The effect of emotion marketing is analyzed by using MATLAB, and the effect of emotion marketing is classified and optimized. Finally, the Chinese word segmentation technology of Chinese Academy of Sciences (CAS) is used to test, the result shows that, The recall rate of text mining considering emotional marketing effect is more than 80%, which can meet the application needs of automobile manufacturers.
【作者单位】: 东华大学旭日工商管理学院;湖北汽车工业学院经济管理学院;北京理工大学网络搜索与挖掘实验室;武汉工商学院;
【基金】:上海市自然科学基金资助项目(15ZR1401600) 湖北省教育厅人文社科指导项目(17G087);湖北省教育厅人文社科资助项目(14Y036) 湖北省高等学校优秀中青年团队计划资助项目(T201411) 武汉工商学院博士启动基金项目(D20160001)
【分类号】:F274;F426.471;H136
,
本文编号:1497255
本文链接:https://www.wllwen.com/gongshangguanlilunwen/1497255.html