Research Of Community Detection And Its Service Push On The Social Q & A Site

With the advent of the era of the Internet for all people,the user community in the social Q&A site is constantly expanding,and more and more questions and answers have been accumulated,forming a dynamic knowledge base.By answering questions and social interaction,social Q&A site has formed the communication mode between users and information,which greatly enriches and satisfies users’ need to acquire and share knowledge,and gradually becomes the favorite knowledge sharing platform for users.Social Q&A site can divide communities of users with similar interests and then push their interested users,questions,answers,etc.,to users in the community in order to improve users’ satisfaction and promote the effective dissemination of knowledge.In addition,Social Q&A site can also help enterprises to promote appropriate ads for different people,which fully reflects its commercial value.Therefore,the research of community detection and its service push on the social Q&A site has important research significance and application value.The main contents of this paper can be divided into four parts.The first part mainly introduces the background and significance of the research,the current research situation at home and abroad,the related theories and techniques.Among them,the related theories include the overview of the social Q&A site,the "Zhihu" data analysis,the graph theory and the social network analysis.The second part is about community detection on the social Q&A site based on link relation.After analyzing the user-linked relationship of the social Q&A site,build a user-linked relationship network based on the user-focused relationship and identify the salient features of the user-linkage network.Finally the algorithm process of Community Detection on the Social Q&A Site Based on Improved Modularity(CDIM)is given.The third part is the research of service push on social Q&A site based on the community detection.This part studies the subject and object,the content and specific way of social Q&A site service push.Based on the analysis of the characteristics of "Zhihu" user behavior,a user interest model is constructed.With the combination of user interest and user-based collaborative filtering,a user cooperative service push algorithm which based on label similarity is proposed.At last,the model and method proposed in this paper are verified by real data sets.Through the experiment,it can be proved that: The similarity calculation formula proposed in this paper is more in line with the characteristics of the user relationship network of social Q&A site;the CDIM algorithm can effectively discover the user relationship network of social Q&A site;the user interest model can effectively represent the user interest;and the effect of personalized service push is better.There are two main innovations in this article: Firstly,community detection on the social Q&A site is based on the users’ following-followed relationship;secondly,user interest is tapped on the basis of community detection to push service on the social Q&A site.