Community Discovery And Commodity Recommendation Model Based On Social Relations Intensity

Online social network is a virtual interpersonal network based on Hypertext Multimedia Link,which is mediated by network and digital symbolic information.With the development and application of information technology in recent years,the degree of differentiation between network and reality is gradually blurred.In real life,people are more dependent on the Internet.Online social networks record the valuable interactive historical data of tens of thousands of social users,and real and quantifiable data resources provide a valuable experimental environment for social computing.The data recorded in online social networking networks implies huge research value and potential business value.Therefore,the excavation and analysis of online social network based on the social network data driven by the user shopping demand is one of the important research contents.This article explores the application of the strength of social relations among friends in the product recommendation through the research and analysis of the online social network.The existing recommendations based on the strength of social relationships mostly focus on friend identification and friend push.Mostly,the recommendation of products in the field of electronic commerce is limited to the recommendation based on personal preferences,neglecting the relationship strength between the friends of social relations and there is recommendation information flooding and other issues,which can not achieve the effective recommendation of product information.In order to better achieve the accurate recommendation of online products,this paper calculates the strength of social network relationships,tap the intensity of the user's social relationships to divide the close relationship with the community,analyze the intimacy between online social relationship users,and propose a social network community discovery model based on the strength of social relations.According to the division of the community,online product recommendation from the individual's preferences extended to similar groups of recommended properties.The main contents of this paper include the following aspects:First,we evaluate the factors that affect the strength of social relationships by analyzing the user's social network data and mining the user's social network content and network structure information.Secondly,by mining the close relationship between community discovery and user relationship strength,a product recommendation model based on strong relational community is constructed.In the process of building the model,we first obtain the behavior data of the user's social network through the data acquisition technology.And then calculate the intensity of social relationships among users according to the frequency of interaction between users and the connection distance between nodes to form a community discovery model that integrates strong social relationships.Finally,the latent vector model of the product to be recommended is solved by decomposing the user behavior matrix in the community.Thirdly,we propose the accuracy of evaluation index of community discovery based on the relationship strength and personalized recommendation.In the area of community discovery,the CDORS method presented in this paper is found to be more efficient than the traditional K-L and LHN methods compared with the evaluation of the modularity and accuracy.The evaluation of the modularity function and accuracy index show that this method can improve the effect of community discovery.According to the experiments such as Recall,MAE and Precision,the recommendation based on strong relational community proposed in this paper is more significant than recommendation based on content and collaborative filtering.