Research On New Energy Vehicle Purchase Decision Based On Sentiment Analysis Of Multi Source Online Reviews

The automobile industry is the pillar industry of the national economy,but the growing demand for automobiles has led to serious environmental pollution and global warming,so the low emission new energy vehicle industry is highly valued by our government.With the popularity and development of the Internet,more and more automobile vertical websites and social media platforms have emerged,providing consumers with channels to share product text reviews.However,due to the large number of online reviews,it is difficult for consumers to make purchase decisions directly based on online product reviews.Therefore,how to help consumers make more ideal purchase decisions based on online text reviews,so as to help carmakers better manufacture,sell vehicles and provide high-quality after-sales service,is an important problem.Guided by this problem,this paper provides a research framework to solve the problem of new energy vehicle product purchase decision-making based on online text reviews in multi vehicle websites.Firstly,under the background of new energy vehicle product selection in the network era,this paper combs and analyzes the characteristics of automobile vertical website,product ranking mode and online reviews.The results show that the existing product ranking method of automobile vertical website can not meet the diversified purchase needs of consumers,so it is reasonable to use online text reviews that can provide emotional description of new energy vehicles to make purchase decisions.Secondly,the new energy vehicle purchase problem is structured as a multiattribute decision-making problem.After preprocessing the multi-source online reviews,the sentiment tendency of each review is analyzed according to the constructed sentiment dictionary of new energy vehicles.The calculated proportion of sentiment tendency is transformed into a probabilistic linguistic term set.Then,considering that the impact of negative reviews is greater than that of positive reviews,the attribute weight is determined based on the statistical method of attribute word frequency of “most satisfied” or “advantage” and “least satisfied” or“disadvantage” review text.The website weight is determined by considering the number and the effectiveness of reviews,and the evaluation information is aggregated into a decision matrix.A multi-attribute decision-making method of TOPSIS,VIKOR and PROMETHEE II based on probabilistic linguistic term set is proposed,and the model fusion is carried out from the perspective of model results.Finally,the proposed new energy vehicle selection decision method is applied to a example of “auto home”,“xcar” and “pcauto”,and the detailed calculation process is shown.Through the analysis of the attributes of each product,and combined with the actual background of the carmaker,the rationality of the proposed new energy vehicle purchase decision-making method is explained,which can provide strong support for the purchase decision of potential consumers,and also help carmakers upgrade and improve products,so as to promote the vigorous development of China’s new energy vehicle industry.