Research On User-Commodity Association Information Interaction System Based On Semantic Similarity

The tremendous enrichment of network information resources and the exponential growth of network users make the problem of how to effectively realize information interaction online an urgent problem for academics and practice.Information interaction simply refers to the process of sending and receiving information,and is the transmission and exchange of information.For e-commerce platform,recommendation and search are the two most basic information interaction scenarios in the system.This thesis constructed a user-commodity association information interaction system model based on semantic similarity for these two information interaction scenarios.Since the proper description of commodity resources and user resources is the premise of effective recommendation and search in the system,the information interaction system designed in this thesis started with the construction of resource description framework,used social tagging under the folksonomy to describe network commodity resources and the network user resources,formed socialized resource description frameworks of individual commodities and socialized resource description frameworks of individual users.On this basis,the association rules between social resource description frameworks are designed.To improve the ability of system to identify and understand resources,this thesis adopted the semantic association method to achieve the docking and matching between commodity resources and user resources.The social tag semantic similarity was calculated between frames of all individual commodity resource description frameworks,while considering the the weight of each social tag to derive the degree of association between all individual commodity resource description frameworks,and based on this,commodity collection resource description frameworks were formed,and the user collection resource description frameworks were formed according to the same method.In the recommended information interaction context,the degree of association between the product collection resource description frameworks and the user collection resource description frameworks was further calculated to achieve two-way recommendation between the product collection and the user collection.In the search information interaction context,the user's search content was processed by word segmentation,the valid words after the word segmentation were regarded as tags to construct the content of the user input framework,and then the degree of association between the user input framework and the user collection or the product collection framework was calculated,and the content of the collection framework with the largest relevance is selected as the search results and finally a relatively complete user-commodity-related information interaction system based on semantic similarity was formed.Finally,this thesis used the social tag resources in Douban reading to conduct a context experiment on the user-commodity-related information interaction system based on semantic similarity,which effectively realizes the docking and matching between network user resources and network commodity resources in the context of recommendation and search interaction.It provides a new method for improving the information interaction effect and efficiency between users and commodities in the field of e-commerce,which has certain theoretical and practical significance.