Topic Recommendation Based On Neighborhood User Model

With the popularity of Internet social networking and the development of social networking,social networking has gradually transformed from a simple communication platform to a main way for people to get information.The social network,because of the number of users and the huge amount of data precipitation,it is difficult for users to get the accurate information they need.In response to this problem,many social network platforms adopt a way of recommending personalized content to users to facilitate users to obtain their own topics of interest.However,due to the existing personalized recommendation research,some of them only consider the user's personal configuration information,and the other part is not deep enough for mining user neighborhood social relations,which will make common recommendation methods unable to locate users accurately.After summarizing the problem of interaction between users in micro-blog social platform's topic recommendation problem,this paper proposes a micro-blog topic recommendation model based on neighborhood users.First,according to the purpose of this paper,we summarize and analyze the literature from three perspectives: user model research,model expansion and personalized recommendation algorithm.The research on the classification,information source,modeling method and updating of personalized recommendation model based on user interest is deeply interpreted.This paper makes a literature survey on the expansion mode of the user model,analyzes the existing problems in the present research results,and lays a theoretical foundation for the next step of research.This paper summarizes the mainstream personalized recommendation algorithms at home and abroad.Secondly,we build the model in this paper,and analyze the user model as the combination of content coverage and professional coverage.The user model is extended based on ontology,and the neighborhood user social relations are added.The relationship between target users and friends is defined as cognitive relationship,including resource cognition and attention relationship cognition,respectively.The interest topic of neighborhood friend set is combined with the target user interest topic set,and the theme set of interest is expanded and the interest degree of topic is updated to generate the final neighborhood user topic set.Finally the use of the Northern Polytechnic neighborhood model set the user data set and micro-blog public data,recommended to verify the effect of the model,results show that the proposed recommendation model based on the theme of the neighborhood user recommendation effect is obviously superior to the isolated user model and collaborative filtering mechanism,and also can meet most people's interest in the theme the user interest model needs through the neighborhood friends set learning that is conducive to enhancing the user social initiative.