Research On A Multi-dimensional Comprehensive Recommendation Method Based On User Relationship Strength In A Social Network Environment

With the rapid development of internet technology,it brings convenience during our lives,but it also brings many negative influences for us.That is,many people are unable to choose the information what they want under the massive data,and the recommendation system can be precisely to solve the overloading information.People usually communicate and share information with other persons through social networking platforms in their daily life,and it is the increasingly significant about the social influences of social network with the continuous growth of the number of users in social networks.Therefore,the analysis of recommendation is based on the influences of the user relationships in the social network,which has become the research direction of many experts in the aspect of recommendation system.At present,the recommendation of social network-based entity is a hot topic in recommendation research.In addition,the recommendation of social network entity is also one of the major issues in the research and analysis of social network.The recommendation system has been studied extensively at present.The traditional recommendation system mainly includes the collaborative filtering recommendation system,the content-based recommendation system etc.Some typical recommendation methods are effectively applied in practical applications.However,there are some problems with the traditional recommendation methods,which are failure to extract more useful information from social network and which do not consider performing the recommendation of entity comprehensively in the view of multiple dimensions.Obviously,this will result in the low accuracy of the recommendation results inevitably.In order to solve the deficiency of the research on traditional recommendation system,this study focuses on the research and analysis of user relationship strength in social network,and proposes a multi-dimensional comprehensive recommendation method which is based on the strength of users' relationship under the environment of social network.The main work of this paper is as follows:(1)The dimensions of the estimation of user relationship strength include the stability of user' reviews,the reliability of user,the frequency of user' interaction,the user' common neighbors and similar communities in the modeling and analysis of user relationship strength;(2)Factors of recommendation mainly include the similarity degree of entity and the user's interest degree in the modeling and analysis of the recommendation factors.Firstly,the definition of the entity similarity and estimating the value of entity similarity which is based on the type of entity,the similarity of the price of entity,the similarity of the quality of the entity and the similarity of the sales of entity;Secondly,the definition of the user interest degree and the corresponding estimation method are given;(3)The implementation of the multi-dimensional comprehensive recommendation algorithm.It mainly includes the algorithm of user candidate set,the algorithm of entity candidate set,the algorithm of user interest degree and the algorithm of comprehensive module for recommendation;(4)Experiments show that the proposed multi-dimensional comprehensive recommendation method is based on user relationship strength under the environment of social network in this study which has better performance than traditional recommendation methods.