The rapid advancement of internet technology has led to an exponential growth in network da-ta.Accurately identifying community structures within this vast and complex pool of network data is of significant importance for in-depth understanding of the network's topological structure,analyzing public opinion on networks,and other research areas.Community structures are a key domain and core feature of complex network research,playing a vital role in uncovering the essential functions of networks.This article systematically reviews the evolution and classification of community detection algorithms,with a special focus on the principles and characteristics of deep learning-based methods such as graph neural networks,graph convolutional neural networks,and autoencoders.It summarizes the research achieve-ments in algorithm optimization and application of such methods and proposes potential research directions for future community detection algorithms.
关键词
社区发现/深度学习/图神经网络/复杂网络
Key words
community detection/deep learning/graph neural network/complex network
维普
万方数据