| With the rapid development of e-commerce industry,customers shopping online is also very popular and frequent,so online sales as an important means of business sales.At present,the e-commerce platform has basically set up a comment system,in which users can express the experience effect on goods.Moreover,the e-commerce platform generates a large number of transaction data and user comment data every day,which contains a lot of valuable information.Therefore,the information mining of consumer's text review data is likely to effectively drive the business growth of merchants,thus providing valuable advice for businesses and platforms to develop new marketing strategies or product improvements.In this paper,we conducted an in-depth analysis of the review data of six brands from Tmall which are Haier,Kowos,Mei,Stone,Felipe and Xiaomi.Through the eight-clawed fish crawler software to crawl the relevant information of the comment data before December 8,2018,and use python software to pre-process the text data,including text de-duplication,short sentence deletion,Chinese participle,filtering deactivation words,and so on.The processed text data was used to analyze the emotional tendency of each brand comment and divides the text comment data into three parts: positive,neutral and negative.Using comments of positive and negative emotions as LDA thematic model analysis,extracts the corresponding theme as well as the high frequency characteristic word,thus carries on the appraisal to the product.At the same time,by means of machine learning and natural language processing,in the aspect of establishing pairwise comparison matrix,the subjective judgment of decision makers will not be relied too much,and the index and weight of participle will be obtained more objectively.Finally,the analytic hierarchy process(AHP)and the fuzzy comprehensive evaluation(FCE)methods were used to analyze the purchase decision of the floor sweeping robot on the e-commerce platform. |
PDF Full Text Request