Emotional Intensity Analysis Of Online Reviews Based On Text Mining

The popularization of the Internet has laid the foundation for the development of e-commerce.The competition among enterprises in the information age has become increasingly fierce.In the past,the idea of "product-centered" was gradually transformed into "customer-centered".Online review of users became a bridge for information communication between buyers and sellers.Users continue to participate in online reviews,expressing their inner feelings can reduce their risk in online shopping in the future,help businesses obtain user feedback,improve product quality and service levels,enhance user stickiness in e-commerce platforms,and increase users' trust in platform content.Based on this,this paper proposes an online comment sentiment analysis and application based on text mining.In this paper,we take the online reviews of M's two e-commerce platforms in Tmall and Jingdong as the research objects.Through the personal experience of M company's products and the information feedback from network users,it is found that there are three major problems in M company,such as low user satisfaction,serious homogeneity of old and unobtrusive product competitive advantage.Combining the research results of scholars at home and abroad with the theoretical basis of user online comments,sentiment analysis,and text mining,a scientific and reasonable solution to these problems is given:First of all,preprocessing the text of M company user commentary data obtained from an ecommerce website,including word segmentation and part-of-speech tagging using the jieba segmentation package in the Python language,so as to obtain a sequence of real words that may contain valid information related to P product reviews.Secondly,Extracting feature pairs from user online reviews by constructing the collocation pattern of feature words and opinion words,and matching them with the sequence of real words,according to the emotional intensity of opinion words,the semantic polarity and intensity of feature words of user reviews are calculated by combining semantic methods and statistical methods;Finally,the user comment feature words are aggregated into product specific feature items,and their emotional intensity values are calculated.According to the results of sentiment analysis,specific suggestions were made on the product design and upgrade of the company and customer service.All of the above is to help companies identify user interests,improve customer satisfaction,clarify the product's competitive advantage and need to upgrade and improve,hoping to provide a number of practical reference solutions for many e-commerce businesses and mobile phone manufacturers.