Research On The Collaborative Filtering Recommendation Method In Mobile Microblog Based On User Interests

With the arrival of the mobile 4G network and the large coverage of the wireless network,the rapid development of web3.0 and the rapid spread of smart phones and tablet PCs,gave birth to microblogging client(mobile micro-blog),and rapid growth in the global development of.Mobile microblogging is a mobile micro-blog service,recommended information for the user requirements,but which is the higher the dependence on the scene.To meet the individual needs of users and reduce the user search information time,the micro-blog site to develop the recommended system.In the process of mobile micro-blog recommended,because of the mobile micro-blog mobility,timeliness characteristics,the traditional recommendation system can't simply copy the application,which need to improve its recommended algorithm.At the same time,the user's personalized attributes,the user's social network relationship and the user's current situation characteristics will have an impact on the user's interest,thus,which have the impact of microblog recommended accuracy in a certain extent.In order to increase the viscosity of the platform and improve the quality of the platform recommendation information,and to meet the individual needs of the mobile micro-blog users at the same time,this paper constructs the user interest model and the social network user interest model by analyzing the interest factors of mobile micro-blog users And finally merges into a comprehensive user interest model,combined with collaborative filtering algorithm,designed a method which based on user interest in mobile micro-blog collaborative filtering recommended.Firstly,the comprehensive user interest model is constructed,and the problem of data sparseness in cooperative filtering recommendation is solved by weighted Slope One method.Secondly,we improve the K-means clustering method,select the user cluster center accurately,improve the accuracy of mobile micro-blog user clustering,and search the nearest neighbor's efficiency.Finally,through the comprehensive user interest model,to obtain accurate target user interest,in the clustering results,through the similarity formula calculation,which generate accurate target user neighbors,predict the target user's accurate recommendation information,so as to provide more mobile microblogging users accurate recommendation information.