Research On The Impact Of Product Characteristics On Consumers’ Online Purchase Intention

With the rapid development of e-commerce,online transaction has become the preferred consumption mode of the masses.Studies proved that the rich product reputation in online shopping platforms can better meet the diverse needs of customers.In order to facilitate consumers to collect product information and improve their shopping experience,major e-commerce platforms have introduced the Customer questions and answers(Customer Q&As)function to provide a platform for asking questions about products to consumers who have already purchased products.As a new type of "User Generated Content",Customer Q&As content has become another important form of online word-of-mouth after the online reviews.However,the existing researches on the mining and analysis of e-commerce products’ Customer Q&As mainly focus on shallow statistics and survey analysis,and most researches on the impact of consumers’ purchase intention only focused on a single dimension,while in reality,consumers tend to consider the characteristics of different aspects of products and then make their purchase decisions.Based on the theory of inquiry learning in the field of education,a multi-dimensional product characteristics index system has been built.In addition,product feature extraction and emotion analysis are carried out on the Customer Q&As text,so as to construct an improved AHP-BP neural network model for analyzing the relationships between product features and consumers’ purchase intention.Empirical analysis is proceed by using the crawling product data to obtain the research results,which provides help for e-commerce platforms and enterprises’ online word-of-mouth operation strategy.The specific research contents of this thesis are as follows:(1)Construction of product characteristics index system.Through literature research and expert consultation,four first-level indicators including online reviews,customer Q&As,product services and stores attribute.In view of the insufficient support of current literature on customer Q&As,crawler software is used to crawl the Customer Q&As texts of Taobao client users,then extracted the themes by LDA model.After that we constructed a customer Q&As emotion dictionary for emotional analysis of the answer texts,so as to obtain the emotional evaluation of products under different theme characteristics.(2)Improve the AHP-BP neural network model.In order to ensure that the research results can not only retain the subjective wills of consumers but also take into account the objective operation of neural network in data fitting.The personality trait theory is used to reduce the subjective influence of experts on the results in the AHP,and initialized the connection weights between input neurons of BP neural network by the index weights obtained from AHP,and an improved AHP-BP neural network model is constructed,according to the final weights of neural network training,the importance degrees of all indicators is obtained.(3)Empirical Study.Satty1-9 scoring method was used to invite experts to score the product characteristics index,and the experts were given the weights by the big Five personality test.After that the weights of each index was obtained based on AHP.Then by crawling the corresponding index data form Taobao products,the adjustment data was obtained with assigning the weight of AHP to the crawling data.Inputting the adjustment data into BP neural network model for network training,and finally the significant importance of each input variable was obtained.In the end,the superiority of the proposed model was verified on the test set.This thesis mined the product characteristic factors which can influence consumers’ purchase intention implied in the customer Q&As,but also constructed the product characteristic index system based on the text analysis results,and by using the improved AHP-BP neural network model to analyze each characteristic index.According to the research results,some marketing suggestions are put forward for e-commerce platforms and enterprises from the following four aspects: online comments,customer questions and answers,product service and store evaluation,and commodity collection.