摘要
为有效指导服装企业生产经营决策,选取服装网购评论为数据样本和研究对象,提出了基于服装网购评论的消费热点情报分析方法,以探究消费者对所采购的服装的关注热点。采用网络爬虫技术采集服装网购评论数据并进行预处理后,利用SnowNLP技术进行情感倾向性分类。在关键词抽取时针对分词过程存在错分专业词汇问题,引入服装专业领域分词词典和停用词典,并结合GooSeeker方法和人工调整方法进一步提高关键词抽取精准度。关键词抽取后建立共词矩阵,并聚类进行社会网络分析得到消费热点情报信息。以真丝服装网购评论进行实证分析以验证方法的有效性。结果发现:真丝服装网购消费者依次易就面料、颜色、尺码、质量等热点关键词给出负面反馈;此外还得到这些热点关键词关联的负面反馈信息及与其他热点关键词之间的相互关系,如面料的负面反馈主要与薄透、褶皱、缩水和引申的丝料价格贵有关,对面料差评的消费者往往会更关注尺码、物流、价格以及退换货等信息。
In order to effectively guide the production and operation decision-making of clothing enterprises,clothing online shopping comments were selected as data samples and research objects.Besides,consumer focus intelligence analysis based on clothing online shopping comments was proposed to explore consumer focus on clothing.After web crawler technology was applied to collect data on clothing online shopping;comments and preprocess them,SnowNLP technology was employed to classify emotion tendencies.During keyword extraction,there was a problem of misclassification of professional vocabulary in the process of word segmentation.Hence,the word segmentation dictionary and out-of-use dictionary in the clothing field were introduced.Meanwhile,GooSeeker method and the manual adjustment method were combined to further improve the accuracy of keyword extraction.After the keywords were extracted,a common word matrix was established and clustered for social network analysis to obtain consumer focus intelligence information.An empirical analysis was carried out for online shopping comments on silk clothing to verify the effectiveness of this method.The results show that,the online shoppers of silk clothing gave negative feedback on such hot keywords as fabric,color,size,and quality.In addition,the negative feedback information associated with these hot keywords and the interaction with other hot keywords were obtained.For example,the negative feedback of fabrics was mainly related to the thinness,wrinkles,shrinkage and the extended price of silk materials.At the same time,consumers who gave bad comments on fabrics tended to pay more attention to size,logistics,price,change and return.
作者
胡觉亮
徐瑶瑶
董建明
HU Jueliang;XU Yaoyao;DONG Jianming(School of Science,Zhejiang Sci-Tech University,Hangzhou 310018,China;School of Fashion Design&Engineering,Zhejiang Sci-Tech University,Hangzhou 310018,China)
出处
《浙江理工大学学报（自然科学版）》
2021年第1期21-30,共10页
Journal of Zhejiang Sci-Tech University(Natural Sciences)
基金
国家自然科学基金项目(11971435)
浙江省自然科学基金项目(LY18A010029)。
关键词
抽取
社会网络分析
Clothing online shopping
comments
consumer focus
classification of emotions
keyword extraction
social network analysis
