Research On Complex Network Community Detection Algorithm Based On Node Relationship

Complex network is a highly abstraction of complex systems in the real world.Exploring the structural characteristics of complex networks is one of the means to recognize and understand complex networks.The community structure of a complex network is a cluster of sparse external associations and tightly connected internals.It is an important method to study the network structure,characteristics and functions by using the community detection algorithm to mine the community structure in the complex network.The research results of the algorithm can also be applied to fields such as information retrieval and protein complex analysis.Although there are a large number of excellent community detection algorithms,they still have certain limitations.Therefore,this paper improves two community detection algorithms,the main content and contributions are as follows:Aiming at the problem of community detection in large-scale networks,this paper proposes to extend the community detection algorithm based on the membership strength of intermediate nodes(EBIAS),which is improved based on the BIAS algorithm.The BIAS algorithm uses the calculation of the membership of the intermediate nodes to divide the communities,which improves the performance of the community structure detection algorithm in large-scale networks.However,the experimental results show that when the community structure is complex,for example,when the LFR generates data with a mixing coefficient greater than 0.5,the performance of the BIAS algorithm is affected and the algorithm is unstable.To solve this problem,this paper introduces the aggregation coefficient and HDI indicators,To further improve the initialization process of the BIAS algorithm and the process of dividing the membership degree of intermediate nodes,thereby improving the efficiency of the detection algorithm for large-scale network communities.Aiming at the problem of protein complex classification,this paper proposes an IFPM algorithm based on the interaction force between protein molecules.The IFPMalgorithm is inspired by the definition of van der Waals force.According to the interaction force between protein molecules,the maximum force of neighboring proteins is selected as the label,and the protein label is updated in the iterative process of label propagation.Through tag propagation,the proteins in the same complex will be assigned the same tag to classify the protein complex.Finally,this paper also uses multiple artificial data sets and real data sets for experiments.The experimental results show that both the EBIAS algorithm and the IFPM algorithm can detect high-quality community structures.