Counterfactual Learning on Graphs:A Survey

作     者：Zhimeng Guo Zongyu Wu Teng Xiao Charu Aggarwal Hui Liu Suhang Wang Zhimeng Guo;Zongyu Wu;Teng Xiao;Charu Aggarwal;Hui Liu;Suhang Wang

作者机构：College of Information Sciences and TechnologyPennsylvania State UniversityUniversity Park 16802USA International Business Machines Corporation T.J.Watson Research CenterNew York 10598USA College of EngineeringMichigan State UniversityEast Lansing 48824USA 

出 版 物：《Machine Intelligence Research》 (机器智能研究(英文版))

年 卷 期：2025年第22卷第1期

页      面：17-59页

核心收录：

学科分类：0711[理学-系统科学] 07[理学] 08[工学] 080203[工学-机械设计及理论] 0802[工学-机械工程] 

基　　金：supported by,or in part by the National Science Foundation(NSF),USA(No.IIS-1909702) Army Research Office(ARO),USA(No.W911NF-21-10198),and Cisco Faculty Research Award 

主　　题：Counterfactual learning graph-structured data graph neural networks fairness explainability 

摘      要：Graph-structured data are pervasive in the real-world such as social networks,molecular graphs and transaction *** neural networks(GNNs)have achieved great success in representation learning on graphs,facilitating various downstream ***,GNNs have several drawbacks such as lacking interpretability,can easily inherit the bias of data and cannot model casual ***,counterfactual learning on graphs has shown promising results in alleviating these *** approaches have been proposed for counterfactual fairness,explainability,link prediction and other applications on *** facilitate the develop-ment of this promising direction,in this survey,we categorize and comprehensively review papers on graph counterfactual *** divide existing methods into four categories based on problems *** each category,we provide background and motivating ex-amples,a general framework summarizing existing works and a detailed review of these *** point out promising future research directions at the intersection of graph-structured data,counterfactual learning,and real-world *** offer a comprehensive view of resources for future studies,we compile a collection of open-source implementations,public datasets,and commonly-used evalu-ation *** survey aims to serve as a“one-stop-shopfor building a unified understanding of graph counterfactual learning cat-egories and current resources.