Demand Analysis Based on E-commerce Platform User Evaluation
This study investigates the feasibility of extracting user demands and levels of concern from user evaluation on e-commerce platforms,which reflect users'attitudes and priorities towards products.A total of 6,348 user reviews for study desk lamps were collected from a prominent domestic e-commerce platform using a Python web crawler.The collect-ed data were then processed using the open-source natural language processing(NLP)tool HANLP,including segmenta-tion,part-of-speech tagging,keyword extraction,word frequency analysis,and word cloud generation.By analyzing the data,user demands and levels of concern were identified in categories such as optical parameters,interaction methods,structure,and product appearance.These insights were further utilized to transform the identified demands into actionable outcomes.The results indicate that this method significantly accelerates the speed of demand analysis when applied to prod-uct research with a substantial amount of user review data.The extracted keywords provide valuable information on user preferences and priorities,enabling product improvement and design optimization.However,for products with limited re-view data,the use of this method may be associated with a certain degree of error.
web spidernatural language processing(NLP)user reviewkeyword extraction
万方数据
维普
