{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,10]],"date-time":"2026-04-10T10:08:19Z","timestamp":1775815699130,"version":"3.50.1"},"reference-count":59,"publisher":"Association for Computing Machinery (ACM)","issue":"6","content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Inf. Syst."],"published-print":{"date-parts":[[2025,11,30]]},"abstract":"<jats:p>\n            Generating\n            <jats:italic toggle=\"yes\">post hoc<\/jats:italic>\n            causal explanations for graph neural network-based recommender systems is vital for enhancing the credibility and interpretability of recommendations. Existing model-agnostic explainers primarily capture statistical correlations between topological information and recommendation outcomes. However, they often fail to identify true causal relationships due to their model-agnostic design and the challenges posed by heterogeneous graph structures. To address these limitations, we propose a causality-inspired graph neural network explainer for recommender systems, namely CaGE, which generates explanations reflecting causality in recommendation scenarios without accessing the internal parameters of the recommender system. Unlike previous explainers that rely on correlation-based learning, CaGE leverages heterogeneous interventional distributions to eliminate backdoor paths of non-causal variables in the structural causal model of the recommendation task, ensuring causation is accurately captured. Specifically, CaGE incorporates backdoor adjustment based on heterogeneous interventional distributions and causal contrastive learning to optimize a set of heterogeneous soft masks that disentangle causation from non-causation. Additionally, a causality-inspired meta-path search strategy is employed to represent causation as paths between users and recommended items, further enhancing explanation readability. Extensive experiments are conducted on three recommendation datasets, and the experimental results illustrate the superior fidelity of CaGE as compared to state-of-the-art baselines.\n          <\/jats:p>","DOI":"10.1145\/3729224","type":"journal-article","created":{"date-parts":[[2025,4,10]],"date-time":"2025-04-10T10:43:11Z","timestamp":1744281791000},"page":"1-29","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":5,"title":["CaGE: A Causality-inspired Graph Neural Network Explainer\u00a0for Recommender Systems"],"prefix":"10.1145","volume":"43","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-1124-9509","authenticated-orcid":false,"given":"Shuo","family":"Yu","sequence":"first","affiliation":[{"name":"Dalian University of Technology, Dalian, China and Key Laboratory of Social Computing and Cognitive Intelligence (Dalian University of Technology), Ministry of Education, Dalian, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-9550-1193","authenticated-orcid":false,"given":"Yicong","family":"Li","sequence":"additional","affiliation":[{"name":"Dalian University of Technology, Dalian, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-6391-375X","authenticated-orcid":false,"given":"Shuo","family":"Wang","sequence":"additional","affiliation":[{"name":"Dalian University of Technology, Dalian, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7356-7196","authenticated-orcid":false,"given":"Tao","family":"Tang","sequence":"additional","affiliation":[{"name":"University of South Australia, Adelaide, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3776-9799","authenticated-orcid":false,"given":"Qiang","family":"Zhang","sequence":"additional","affiliation":[{"name":"Dalian University of Technology, Dalian, China and Key Laboratory of Social Computing and Cognitive Intelligence (Dalian University of Technology), Ministry of Education, Dalian, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3373-6164","authenticated-orcid":false,"given":"Jingjing","family":"Zhou","sequence":"additional","affiliation":[{"name":"Zhejiang Gongshang University, Hangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2826-6367","authenticated-orcid":false,"given":"Ivan","family":"Lee","sequence":"additional","affiliation":[{"name":"University of South Australia, Adelaide, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8324-1859","authenticated-orcid":false,"given":"Feng","family":"Xia","sequence":"additional","affiliation":[{"name":"RMIT University, Melbourne, Australia"}]}],"member":"320","published-online":{"date-parts":[[2025,9,11]]},"reference":[{"key":"e_1_3_1_2_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-40501-3_45"},{"key":"e_1_3_1_3_2","first-page":"6486","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Chen Xuexin","year":"2024","unstructured":"Xuexin Chen, Ruichu Cai, Zhengting Huang, Yuxuan Zhu, Julien Horwood, Zhifeng Hao, Zijian Li, and Jos\u00e9 Miguel Hern\u00e1ndez-Lobato. 2024. Feature attribution with necessity and sufficiency via dual-stage perturbation test for causal explanation. In Proceedings of the International Conference on Machine Learning, 6486\u20136502."},{"key":"e_1_3_1_4_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2020.3008732"},{"key":"e_1_3_1_5_2","doi-asserted-by":"publisher","DOI":"10.1145\/3442381.3449940"},{"key":"e_1_3_1_6_2","doi-asserted-by":"publisher","DOI":"10.1145\/3397271.3401051"},{"key":"e_1_3_1_7_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2022.3201170"},{"key":"e_1_3_1_8_2","doi-asserted-by":"publisher","DOI":"10.1145\/3639048"},{"key":"e_1_3_1_9_2","doi-asserted-by":"publisher","DOI":"10.1145\/3336191.3371824"},{"key":"e_1_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1145\/3477495.3531828"},{"key":"e_1_3_1_11_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2022.108601"},{"key":"e_1_3_1_12_2","doi-asserted-by":"publisher","DOI":"10.1145\/3583780.3614909"},{"key":"e_1_3_1_13_2","doi-asserted-by":"publisher","DOI":"10.1145\/3580488"},{"key":"e_1_3_1_14_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2022.3185388"},{"key":"e_1_3_1_15_2","doi-asserted-by":"publisher","DOI":"10.1145\/3579993"},{"key":"e_1_3_1_16_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v37i7.26040"},{"key":"e_1_3_1_17_2","first-page":"1","volume-title":"Proceedings of the International Conference on Learning Representations","author":"Li Wenqian","year":"2023","unstructured":"Wenqian Li, Yinchuan Li, Zhigang Li, Jianye\u00a0 HAO, and Yan Pang. 2023. DAG matters! GFlowNets enhanced explainer for graph neural networks. In Proceedings of the International Conference on Learning Representations, 1\u201325."},{"key":"e_1_3_1_18_2","doi-asserted-by":"publisher","DOI":"10.1145\/3637528.3671782"},{"key":"e_1_3_1_19_2","first-page":"6666","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Lin Wanyu","year":"2021","unstructured":"Wanyu Lin, Hao Lan, and Baochun Li. 2021. Generative causal explanations for graph neural networks. In Proceedings of the International Conference on Machine Learning, 6666\u20136679."},{"key":"e_1_3_1_20_2","doi-asserted-by":"publisher","DOI":"10.1109\/TNSE.2021.3110677"},{"key":"e_1_3_1_21_2","doi-asserted-by":"publisher","DOI":"10.1145\/3616855.3635855"},{"key":"e_1_3_1_22_2","first-page":"19620","volume-title":"Proceedings of the Neural Information Processing Systems","author":"Luo Dongsheng","year":"2020","unstructured":"Dongsheng Luo, Wei Cheng, Dongkuan Xu, Wenchao Yu, Bo Zong, Haifeng Chen, and Xiang Zhang. 2020. Parameterized explainer for graph neural network. In Proceedings of the Neural Information Processing Systems, 19620\u201319631."},{"key":"e_1_3_1_23_2","doi-asserted-by":"publisher","DOI":"10.1145\/3308558.3313607"},{"issue":"2","key":"e_1_3_1_24_2","first-page":"28","article-title":"Explainable meta-path based recommender systems","volume":"3","author":"Markchom Thanet","year":"2024","unstructured":"Thanet Markchom, Huizhi Liang, and James Ferryman. 2024. Explainable meta-path based recommender systems. ACM Transactions on Recommender Systems 3, 2 (2024), 28 pages.","journal-title":"ACM Transactions on Recommender Systems"},{"key":"e_1_3_1_25_2","unstructured":"Thanet Markchom Huizhi Liang and James Ferryman. 2024. Review of explainable graph-based recommender systems. arXiv:2408.00166. Retrieved from https:\/\/arxiv.org\/abs\/2408.00166"},{"key":"e_1_3_1_26_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.01103"},{"key":"e_1_3_1_27_2","doi-asserted-by":"publisher","DOI":"10.1145\/3543507.3583374"},{"key":"e_1_3_1_28_2","doi-asserted-by":"publisher","DOI":"10.1145\/3485447.3511951"},{"key":"e_1_3_1_29_2","doi-asserted-by":"publisher","DOI":"10.1145\/3459637.3482420"},{"key":"e_1_3_1_30_2","doi-asserted-by":"publisher","DOI":"10.1145\/3694682"},{"issue":"1","key":"e_1_3_1_31_2","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3592853","article-title":"Multi-view enhanced graph attention network for session-based music recommendation","volume":"42","author":"Wang Dongjing","year":"2023","unstructured":"Dongjing Wang, Xin Zhang, Yuyu Yin, Dongjin Yu, Guandong Xu, and Shuiguang Deng. 2023. Multi-view enhanced graph attention network for session-based music recommendation. ACM Transactions on Information Systems 42, 1 (2023), 1\u201330.","journal-title":"ACM Transactions on Information Systems"},{"key":"e_1_3_1_32_2","doi-asserted-by":"publisher","DOI":"10.1145\/3627826"},{"key":"e_1_3_1_33_2","first-page":"1","volume-title":"Proceedings of the International Conference on Learning Representations","author":"Wang Xiaoqi","year":"2023","unstructured":"Xiaoqi Wang and Han Wei Shen. 2023. GNNInterpreter: A probabilistic generative model-level explanation for graph neural networks. In Proceedings of the International Conference on Learning Representations, 1\u201325."},{"key":"e_1_3_1_34_2","first-page":"18446","volume-title":"Proceedings of the Neural Information Processing Systems","volume":"34","author":"Wang Xiang","year":"2021","unstructured":"Xiang Wang, Yingxin Wu, An Zhang, Xiangnan He, and Tat-Seng Chua. 2021. Towards multi-grained explainability for graph neural networks. In Proceedings of the Neural Information Processing Systems, Vol. 34, 18446\u201318458."},{"issue":"3","key":"e_1_3_1_35_2","first-page":"2229","article-title":"Time-aware service recommendation with social-powered graph hierarchical attention network","volume":"16","author":"Wei Chunyu","year":"2023","unstructured":"Chunyu Wei, Yushun Fan, and Jia Zhang. 2023. Time-aware service recommendation with social-powered graph hierarchical attention network. IEEE Transactions on Services Computing 16, 3 (2023), 2229\u20132240.","journal-title":"IEEE Transactions on Services Computing"},{"key":"e_1_3_1_36_2","doi-asserted-by":"publisher","DOI":"10.1145\/3593313"},{"key":"e_1_3_1_37_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2020.3048414"},{"key":"e_1_3_1_38_2","doi-asserted-by":"publisher","DOI":"10.1145\/3331184.3331203"},{"key":"e_1_3_1_39_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICDM50108.2020.00079"},{"issue":"1","key":"e_1_3_1_40_2","first-page":"1","article-title":"A reusable model-agnostic framework for faithfully explainable recommendation and system scrutability","volume":"42","author":"Xu Zhichao","year":"2023","unstructured":"Zhichao Xu, Hansi Zeng, Juntao Tan, Zuohui Fu, Yongfeng Zhang, and Qingyao Ai. 2023. A reusable model-agnostic framework for faithfully explainable recommendation and system scrutability. ACM Transactions on Information Systems 42, 1 (2023), 1\u201329.","journal-title":"ACM Transactions on Information Systems"},{"key":"e_1_3_1_41_2","doi-asserted-by":"publisher","DOI":"10.1145\/3485447.3512273"},{"key":"e_1_3_1_42_2","doi-asserted-by":"publisher","DOI":"10.1145\/3580305.3599289"},{"key":"e_1_3_1_43_2","first-page":"9240","volume-title":"Proceedings of the Neural Information Processing Systems","author":"Ying Zhitao","year":"2019","unstructured":"Zhitao Ying, Dylan Bourgeois, Jiaxuan You, Marinka Zitnik, and Jure Leskovec. 2019. Gnnexplainer: Generating explanations for graph neural networks. In Proceedings of the Neural Information Processing Systems, 9240."},{"key":"e_1_3_1_44_2","first-page":"2084","volume-title":"Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining","author":"Yu Junliang","unstructured":"Junliang Yu, Hongzhi Yin, Min Gao, Xin Xia, Xiangliang Zhang, and Nguyen Quoc Viet Hung. 2021. Socially-aware self-supervised tri-training for recommendation. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, 2084\u20132092."},{"key":"e_1_3_1_45_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2020.3033673"},{"key":"e_1_3_1_46_2","first-page":"1","article-title":"Ke Sun, and Honglong Chen. 2024. Formulating and representing multiagent systems with hypergraphs","author":"Yu Shuo","year":"2024","unstructured":"Shuo Yu, Huafei Huang, Yanming Shen, Pengfei Wang, Qiang Zhang, Ke Sun, and Honglong Chen. 2024. Formulating and representing multiagent systems with hypergraphs. IEEE Transactions on Neural Networks and Learning Systems 36, 3 (2024), 1\u201315.","journal-title":"IEEE Transactions on Neural Networks and Learning Systems"},{"key":"e_1_3_1_47_2","first-page":"12241","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Yuan Hao","year":"2021","unstructured":"Hao Yuan, Haiyang Yu, Jie Wang, Kang Li, and Shuiwang Ji. 2021. On explainability of graph neural networks via subgraph explorations. In Proceedings of the International Conference on Machine Learning, 12241\u201312252."},{"key":"e_1_3_1_48_2","doi-asserted-by":"publisher","DOI":"10.1145\/3637528.3672354"},{"key":"e_1_3_1_49_2","doi-asserted-by":"publisher","DOI":"10.1145\/3580305.3599435"},{"key":"e_1_3_1_50_2","first-page":"19810","volume-title":"Proceedings of the Neural Information Processing Systems","volume":"35","author":"Zhang Shichang","year":"2022","unstructured":"Shichang Zhang, Yozen Liu, Neil Shah, and Yizhou Sun. 2022. Gstarx: Explaining graph neural networks with structure-aware cooperative games. In Proceedings of the Neural Information Processing Systems, Vol. 35, 19810\u201319823."},{"key":"e_1_3_1_51_2","doi-asserted-by":"publisher","DOI":"10.1145\/3158369"},{"key":"e_1_3_1_52_2","first-page":"3784","volume-title":"Proceedings of the ACM Web Conference","author":"Zhang Shichang","unstructured":"Shichang Zhang, Jiani Zhang, Xiang Song, Soji Adeshina, Da Zheng, Christos Faloutsos, and Yizhou Sun. 2023. PaGE-Link: Path-based graph neural network explanation for heterogeneous link prediction. In Proceedings of the ACM Web Conference, 3784\u20133793."},{"key":"e_1_3_1_53_2","first-page":"1","article-title":"Feature-enhanced neural collaborative reasoning for explainable recommendation","author":"Zhang Xiaoyu","year":"2024","unstructured":"Xiaoyu Zhang, Shaoyun Shi, Yishan Li, Weizhi Ma, Peijie Sun, and Min Zhang. 2024. Feature-enhanced neural collaborative reasoning for explainable recommendation. ACM Transactions on Information Systems (2024), 1\u201332.","journal-title":"ACM Transactions on Information Systems"},{"key":"e_1_3_1_54_2","doi-asserted-by":"publisher","DOI":"10.1561\/1500000066"},{"key":"e_1_3_1_55_2","doi-asserted-by":"publisher","DOI":"10.1145\/3665325"},{"key":"e_1_3_1_56_2","doi-asserted-by":"publisher","DOI":"10.1145\/3469799"},{"key":"e_1_3_1_57_2","doi-asserted-by":"publisher","DOI":"10.1145\/3477495.3531895"},{"key":"e_1_3_1_58_2","doi-asserted-by":"publisher","DOI":"10.1145\/3485447.3512098"},{"key":"e_1_3_1_59_2","doi-asserted-by":"publisher","DOI":"10.1145\/3653985"},{"key":"e_1_3_1_60_2","doi-asserted-by":"publisher","DOI":"10.1145\/3663760"}],"container-title":["ACM Transactions on Information Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3729224","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,11]],"date-time":"2025-09-11T20:57:18Z","timestamp":1757624238000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3729224"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9,11]]},"references-count":59,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2025,11,30]]}},"alternative-id":["10.1145\/3729224"],"URL":"https:\/\/doi.org\/10.1145\/3729224","relation":{},"ISSN":["1046-8188","1558-2868"],"issn-type":[{"value":"1046-8188","type":"print"},{"value":"1558-2868","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9,11]]},"assertion":[{"value":"2024-10-02","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-03-28","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-09-11","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}