摘要
本文针对北大荒米业竞争对手的部分产品,进行评论数据的文本挖掘与情感分析。为了保证评论数据文本挖掘的质量和全面性,本文利用网络爬虫分别从天猫超市和京东平台爬取了评论数据,接着对爬取到的评论数据进行预处理——文本去重、中文分词、停用词过滤等,再利用R语言对其数据集进行可视化分析。然后,本文基于LDA模型,提取出评论文本中包含的主题词。最后,利用情感词典,进行情感倾向标注,得到在不同主题词下的消费者满意度,提炼出各个品牌的差异化买点,为北大荒米业改进产品、制定合适的销售策略提供决策依据。
In this paper, we mainly discuss the text mining andemotion analysis of the competitors＇s products for BeiDaHuangRice Industry. In order to ensure quality and comprehensive oftext mining, we crawl the product reviews from both Tmall andJingDong using web crawler. Firstly, we preprocess the datas——text repetition reducing, Chinese word segmentation, stopwords filtering and so on. Besides, we use R language to makevisual analysis of the data set. Secondly, based on LDA model,the subject words contained in the comment text are extracted.Finally, using the emotional dictionary to make emotional labeling, getting customer satisfaction under different subjectheadings, refining the differentiation selling points of eachbrand and providing decision basis for the appropriate salesstrategy of BeiDaHuang Rice Industry.
作者
安翔
李世鑫
白雪
杜禹墨
An Xiang;Li Shixin;Bai Xue;Du Yumo(College of Basic Science;Harbin University of Commerc;Harbi;15002)
出处
《北方经贸》
2018年第8期44-47,共4页
Northern Economy and Trade
基金
哈尔滨商业大学大学生创新创业项目(201710240110)
