Research On Compact Subgraph Discovery And Information Recommendation Algorithms In Social Networks

With the rapid development of e-commerce and social tools,the scale and diversity of social networks continue to expand.The research of social networks is becoming a hot topic in the research of data mining.It is of great commercial value and practical significance to acquire close groups in social networks,such as placing advertisements for specific groups of people in social networks and tapping partners with close relationships in customer relationships.The influence of social networks mainly researches on the phenomenon of individuals or groups in social networks,their thoughts or actions interacting with each other.Social network influence analysis has important implications for understanding how information is spread in social networks,how sensation develops,how ideas are innovated,and how experiences are taught.With the diversification of social tools and social methods,we have witnessed many different social ways,such as video socialization,live social networking and so on.Analyzing and mining the influence of social networks can provide new ideas for how individuals in social networks to groups and the role of groups and individuals,and how different information to spread in different social networks.For this reason,this paper proposes a close subgraph discovery framework basing on influence,and analyzes social networks based on the topological structure of social networks to discover the tight subgraphs in social networks.Firstly,the weight of the quantized vertex is the similarity between the vertices.Secondly,by changing the average distribution weights by the similarity between vertices,the improved PageRank algorithm calculates the potential influence of each vertex.Finally,by setting the threshold,quantify the influence scores between the two points to find a tight subgraph in the social network.The experimental results show that the influencing-based close subgraph discovery algorithm not only shows a good balance between the calculation of the personal influence and the common influence among the member vertices,but also applies to the real social network.For the close subgraph based on influence calculation,this paper also studies information recommendation based on feedback.By giving feedback on items that have been recommended,the weight and effectiveness of each member of the social network can be constantly adjusted,thereby improving the accuracy of the recommendation.