Research On Product Portrait Construction Based On Knowledge Graph

With the popularization and normalization of online shopping,the product data and comment data on various e-commerce platforms are increasing day by day.In this context,all the shopping platforms have faced a difficult problem:how to effectively organize,store and mine the huge product data resources accumulated by every platform,in order to improve consumers' online shopping experience and efficiency.The existing e-commerce platforms have realized the organization and display of product static description information,but there are still some shortcomings in mining unstructured product reviews and combining static product information with dynamic product reviews.From the perspective of application requirements,this paper redefined the concept of product portrait,and introduced the theory and method of knowledge graph to study how to construct product portrait.Knowledge graph is an advanced technology that can be used to organize,store and manage massive structure information.Applying it to the construction of product portrait will organize and mine static product information and dynamic product reviews effectively,so as to help e-commerce platforms in product contrast,product search and so on,ultimately provide better product service for users.In this paper,we first defined the product portrait on the basis of user portrait and other related research.Combining with the knowledge graph features,we analyzed the feasibility of constructing product portrait,and on this basis,we put forward the overall framework of the research.Secondly,we analyzed the application requirements of the e-commerce platforms based on the concept of product portrait,and further analyzed the portrait's dimensions and their relevance according to consumer surplus theory and consumer satisfaction theory.Thirdly,the logical structure and technical processes of knowledge graph were introduced and the construction methods of product portrait were designed.Specifically,the logical structure was divided into pattern layer and data layer,and the construction processes were divided into data acquisition,data extraction,data fusion,storage and display.Finally,we took the mobile phone product for example.According to our proposed methods,the portrait of mobile phone product was generated and its application scene was discussed.The research confirms that the concept of product portrait has high application value.Constructing product portrait by knowledge graph we proposed is feasible and effective,and the existing product contrast mechanism on e-commerce platforms also can be improved based on the product portrait.