Semi-supervised community detection method based on generative adversarial networks

作者全名：Liu, Xiaoyang; Zhang, Mengyao; Liu, Yanfei; Liu, Chao; Li, Chaorong; Wang, Wei; Zhang, Xiaoqin; Bouyer, Asgarali

作者地址：[Liu, Xiaoyang; Liu, Chao; Li, Chaorong] Chongqing Univ Technol, Chongqing, Peoples R China; [Zhang, Mengyao] Chongqing Metropolitan Coll Sci & Technol, Sch Artificial Intelligence & Big Data, Chongqing 402160, Peoples R China; [Liu, Yanfei] Tianjin Univ, Coll Intelligence & Comp, Tianjin 300000, Peoples R China; [Wang, Wei] Chongqing Med Univ, Sch Publ Hlth, Chongqing 400016, Peoples R China; [Zhang, Xiaoqin] Chongqing Commun Design Inst Co Ltd, Chongqing 400041, Peoples R China; [Bouyer, Asgarali] Azarbaijan Shahid Madani Univ, Fac Comp Engn & Informat Technol, Tabriz, Iran; [Bouyer, Asgarali] Istinye Univ, Fac Engn & Nat Sci, Dept Software Engn, Istanbul, Turkiye

通信作者：Liu, XY (通讯作者)，Chongqing Univ Technol, Chongqing, Peoples R China.

来源：JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES

ESI学科分类：　

WOS号：WOS:001218218400001

JCR分区：Q1

影响因子：5.2

年份：2024

卷号：36

期号：3

开始页：　

结束页：　

文献类型：Article

关键词：Generative adversarial networks; Community detection; Semi-unsupervised learning; Complex networks

摘要：Community detection in complex networks often suffers from insufficient data and limited utilization of prior knowledge. In this paper we propose "Semi-supervised Generative Adversarial Network" (GANSE), a novel algorithm that integrates Generative Adversarial Networks (GANs) and semi-supervised learning to address these challenges. This method addresses the issues above through a multi-step process. Initially, the network is rewired using vertex similarity metrics, thereby enhancing its structural integrity. Subsequently, a novel generative adversarial network model is designed, and our model facilitates the reconstruction of the network, thereby yielding partitions. Which form the basis for identifying core communities. Additionally, the local clustering coefficient is incorporated as a reward signal and injected into the node selection process. Moreover, isolated nodes are reallocated, ultimately culminating in the derivation of the final community structure. Experimental results on four large real-life datasets demonstrate the clear superiority of the proposed algorithm in terms of F1 and Jaccard metrics when compared to existing algorithms. Notably, our GANSE method outperforms the traditional algorithms in networks with "missing data". Thus showing its robustness and effectiveness in realworld incomplete datasets. Our findings highlight the potential of GANs and semi-supervised learning for enhancing community detection accuracy in complex networks.

基金机构：Chongqing Federation of Social Sciences Key Project [2023NDZD09]; Humanities and Social Sciences Research Key Project of Chongqing Municipal Education Commission [23SKGH247]

基金资助正文：This work is supported in part by Chongqing Federation of Social Sciences Key Project (2023NDZD09) , Humanities and Social Sciences Research Key Project of Chongqing Municipal Education Commission (23SKGH247) .