Research On Complex Network Community Algorithm And Its Information Identification Application

Complex networks has attracted the attention of many scholars since it was proposed.After years of development,the application value of complex network gradually shows up,and has been recognized by the society.Complex network can be applied in many fields,such as analyzing the social relationship between people,analyzing the urban traffic network lines and so on.Driven by many experts and scholars,complex networks have been applied in many fields,such as biology,sociology,computer science,management,and even language and literature.Community detection algorithm is an important subject in complex networks,and social network is a typical complex network.Through the complex network community algorithm to mine and analyze the useful information in social network,it can highlight the application value of community algorithm.However,the emergence of rumors on social networks makes many users obtain untrue information.A large amount of false information has emerged endlessly on social networks.The characteristics of social networks also make rumors spread rapidly on the network,which is difficult to contain.At present,most of rumor detection methods are based on machine learning,and they use the content features of the text,without taking into account the community structure information in social networks.In order to better detect rumors on social networks,this paper proposes RICD(Rumor Identification with Community Detection)framework.By capturing Weibo data on Sina Weibo social platform,the official labeled information is used to extract rumor features and analyzes them.Based on the features of Weibo content and the feature set based on user network,machine learning algorithms and community algorithms are connected for the first time.Combining the features of Weibo content with the features based on user relations newly proposed in this paper,the rumor detection results are optimized,which provides a new idea for rumor detection.Finding the source of rumor can prevent rumor from spreading.This paper summarizes the factors that affect the rumor detection in the network by researching the technology of rumor detection.Based on the rumor detection algorithm MPA(Message-Passing Algorithm),an improved algorithm IMPA(Improved MessagePassing Algorithm)is proposed,which improves the accuracy and speeds up execution efficiency of MPA.In the experimental part,the IMPA algorithm is verified by four large-scale complex network data sets,and the experimental results show that the IMPA is slightly better than the MPA algorithm.