Community Detection in Complex Overlapping Networks Using Graph Autoencoders with Semi-supervised Fuzzy Clustering

This chapter delves into the challenges of community detection in complex networks with overlapping structures, focusing on the limitations of traditional methods and the need for more advanced techniques. It introduces a novel framework that combines graph autoencoders (GAEs) with semi-supervised fuzzy clustering algorithms, specifically Semi-Supervised Fuzzy C-Means (sSFCM) and Semi-Supervised Fuzzy C-Means with Multiple Fuzzification Coefficients (sSMC-FCM). The framework aims to leverage the strengths of both graph representation learning and fuzzy clustering to achieve more accurate and interpretable community detection. The chapter provides a detailed overview of the proposed method, including the feature representation learning process, supervised set construction, and community detection steps. It also presents experimental results on five benchmark datasets, comparing the performance of the proposed method against the well-established GCNFCM baseline. The results demonstrate significant improvements in various evaluation metrics, highlighting the effectiveness of the proposed approach. The chapter concludes with a discussion on the implications of the findings and potential future directions for research in this field.