Opinion Leader Recognition And Friend Recommendation Algorithm Based On Overlapping Communities In Social Networks

With the rapid development of information technology and the popularization of smart devices,all kinds of social networks have become the most extensive virtual platform that affects the lives of people.The work,learning and life patterns that people used to take for granted have already occurred with the development of technology.Earth-shaking changes.Due to the variety of nodes and intricate relationships that exist in the network,it is often difficult to imagine how to find effective information in a real social network and apply it reasonably.A friend recommendation algorithm that can realize opinion leader-led opinion leader identification and solve information overload problems has always been the focus of social network research.In response to these two hot issues,the main research work of this article is as follows:(1)This paper proposes a three-dimensional feature analysis opinion leader recognition algorithm(TDF)based on overlapping communities.In order to improve the accuracy of the detection results,this paper first uses the overlapping community discovery algorithm to divide the entire network into communities,reduce the complexity of the network structure,and improve the efficiency of feature analysis.Then through the analysis of the structural characteristics,behavioral characteristics and emotional characteristics of the nodes,the opinion leader recognition process is more in line with the structural characteristics of the real social network.Finally,the node's influence value is calculated according to the results of node feature analysis,and the node with a relatively larger-node influence value is put into the seed set as the final result to complete the opinion leader identification.The algorithm was verified on four different real social network data sets.The experimental results show that the algorithm is effective and can accurately identify opinion leaders in the network.(2)This paper proposes a dual strategy analysis friend recommendation algorithm(DMF)based on meta path.In order to solve the problem that most of the current friend recommendation algorithms are based on the idea of common friends and the collaborative filtering ideas of a single dimension,the accuracy of the recommendation results is not high enough.This paper proposes a friend recommendation algorithm based on meta-path dual strategy analysis.The algorithm s the real social network into a binary heterogeneous network,and then uses the instance path to calculate the weight of the meta path based on the idea of the meta path,and introduces the node influence value calculated by the TDF algorithm when calculating the final connection probability.And user activity within a certain period of time to improve the accuracy of recommendations.Experiments were conducted on four different real social network data sets,and three friend recommendation algorithms based on different ideas were compared to verify that the DMF algorithm recommendation results have improved accuracy,recall and F indicators.