Communication Network Behavior Feature Mining Based on Community Detection

With the continuous development of information technology and the expansion of communication demand, data in communication network is experiencing explosive growth. Mining communication behavior characteristics in large-scale communication data and capturing nodes with similar communication behaviors and patterns can enhance the understanding of communication traffic, which is crucial for analyzing network communication patterns, managing security, and optimizing networks. In a communication network, when nodes engage in network services and tasks, the complex network often exhibits community characteristics, wherein nodes with the same communication behavior and close connections typically belong to the same community. To address this issue, this paper proposes a method for mining IP network communication behavior features based on community detection. The method first constructs a communication similarity matrix using the communication relationship and frequency between nodes. Then, it utilizes non-negative matrix factorization framework to detect communities in the network, thereby grouping nodes with similar communication patterns into the same community. The results of community division can serve as a foundation for decision-making and subsequent tasks such as network optimization.