{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,13]],"date-time":"2026-11-13T12:37:52Z","timestamp":1794573472267,"version":"4.9.1"},"reference-count":39,"publisher":"Association for Computing Machinery (ACM)","issue":"5","license":[{"start":{"date-parts":[[2026,5,23]],"date-time":"2026-05-23T00:00:00Z","timestamp":1779494400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"funder":[{"DOI":"10.13039\/501100000923","name":"Australian Research Council","doi-asserted-by":"crossref","award":["FT210100624"],"award-info":[{"award-number":["FT210100624"]}],"id":[{"id":"10.13039\/501100000923","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Discovery Projects","award":["DP240101108 and DP260100326"],"award-info":[{"award-number":["DP240101108 and DP260100326"]}]},{"name":"Linkage Projects","award":["LP230200892 and LP240200546"],"award-info":[{"award-number":["LP230200892 and LP240200546"]}]},{"name":"HKU-SCF FinTech Academy, the Shenzhen-Hong Kong-Macao Science and Technology Plan Project","award":["SGDX20210823103537030"],"award-info":[{"award-number":["SGDX20210823103537030"]}]},{"name":"Theme-based Research Scheme of the RGC, Hong Kong","award":["T35-710\/20-R"],"award-info":[{"award-number":["T35-710\/20-R"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Inf. Syst."],"published-print":{"date-parts":[[2026,6,30]]},"abstract":"<jats:p>\n                    Sequential recommendation systems require both\n                    <jats:italic toggle=\"yes\">temporal efficiency<\/jats:italic>\n                    to handle long interaction histories and\n                    <jats:italic toggle=\"yes\">hierarchical representation<\/jats:italic>\n                    to model complex user\u2013item relationships. Existing approaches face a fundamental tension: Mamba-based methods offer linear-time efficiency (\n                    <jats:inline-formula content-type=\"math\/tex\">\n                      <jats:tex-math notation=\"LaTeX\" version=\"MathJax\">\\(\\mathcal{O}(L)\\)<\/jats:tex-math>\n                    <\/jats:inline-formula>\n                    ) but operate in Euclidean space, which distorts hierarchical patterns; hyperbolic models capture taxonomies well but suffer quadratic complexity (\n                    <jats:inline-formula content-type=\"math\/tex\">\n                      <jats:tex-math notation=\"LaTeX\" version=\"MathJax\">\\(\\mathcal{O}(L^{2})\\)<\/jats:tex-math>\n                    <\/jats:inline-formula>\n                    ). To solve this\n                    <jats:italic toggle=\"yes\">dual challenge<\/jats:italic>\n                    , we propose Hyperbolic Mamba (HMamba), the first architecture that\n                    <jats:italic toggle=\"yes\">unifies curvature-aware state spaces<\/jats:italic>\n                    with hyperbolic geometry. Our key insight is that hyperbolic curvature\n                    <jats:inline-formula content-type=\"math\/tex\">\n                      <jats:tex-math notation=\"LaTeX\" version=\"MathJax\">\\(\\kappa\\)<\/jats:tex-math>\n                    <\/jats:inline-formula>\n                    simultaneously governs: (1) state transition granularity through\n                    <jats:inline-formula content-type=\"math\/tex\">\n                      <jats:tex-math notation=\"LaTeX\" version=\"MathJax\">\\(\\mathbf{\\bar{A}}=\\exp(\\Delta\\mathbf{A}\\odot\\mathbf{K}(\\kappa))\\)<\/jats:tex-math>\n                    <\/jats:inline-formula>\n                    and (2) hierarchical distance preservation via\n                    <jats:inline-formula content-type=\"math\/tex\">\n                      <jats:tex-math notation=\"LaTeX\" version=\"MathJax\">\\(d_{\\mathcal{L}}\\propto\\sqrt{\\kappa}\\log(\\cdot)\\)<\/jats:tex-math>\n                    <\/jats:inline-formula>\n                    . This enables joint optimization of efficiency and hierarchy\u2014addressing the previously unsolved problem of\n                    <jats:italic toggle=\"yes\">deep-long modeling<\/jats:italic>\n                    . Experiments show HMamba achieves 3\u201311% accuracy gains while maintaining 3.2\n                    <jats:inline-formula content-type=\"math\/tex\">\n                      <jats:tex-math notation=\"LaTeX\" version=\"MathJax\">\\(\\times\\)<\/jats:tex-math>\n                    <\/jats:inline-formula>\n                    faster training than attention-based models, establishing a new paradigm for hierarchy-aware sequential recommendation. The code and datasets accompanying our paper are publicly available at\n                    <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" ext-link-type=\"uri\" xlink:href=\"https:\/\/github.com\/CoderPowerBeyond\/HMamba\">https:\/\/github.com\/CoderPowerBeyond\/HMamba<\/jats:ext-link>\n                    .\n                  <\/jats:p>","DOI":"10.1145\/3811405","type":"journal-article","created":{"date-parts":[[2026,4,21]],"date-time":"2026-04-21T14:29:20Z","timestamp":1776781760000},"page":"1-27","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":5,"title":["<scp>HMamba<\/scp>\n                    : Hyperbolic Mamba for Sequential Recommendation"],"prefix":"10.1145","volume":"44","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5843-6187","authenticated-orcid":false,"given":"Qianru","family":"Zhang","sequence":"first","affiliation":[{"name":"School of Computing and Data Science, The University of Hong Kong, Hong Kong, Hong Kong"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1830-5932","authenticated-orcid":false,"given":"Honggang","family":"Wen","sequence":"additional","affiliation":[{"name":"School of Computing and Data Science, The University of Hong Kong, Hong Kong, Hong Kong"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9400-842X","authenticated-orcid":false,"given":"Wei","family":"Yuan","sequence":"additional","affiliation":[{"name":"School of Electrical Engineering and Computer Science, The University of Queensland, Brisbane, Queensland, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5227-7210","authenticated-orcid":false,"given":"Crystal","family":"Chen","sequence":"additional","affiliation":[{"name":"School of Computer Science, Boston University, Boston, Massachusetts, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2510-5282","authenticated-orcid":false,"given":"Menglin","family":"Yang","sequence":"additional","affiliation":[{"name":"School of Computer Science, Hong Kong University of Science and Technology, Hong Kong, Hong Kong"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3975-8500","authenticated-orcid":false,"given":"Siu-Ming","family":"Yiu","sequence":"additional","affiliation":[{"name":"School of Computer Science, The University of Hong Kong, Hong Kong, Hong Kong"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1395-261X","authenticated-orcid":false,"given":"Hongzhi","family":"Yin","sequence":"additional","affiliation":[{"name":"School of Electrical Engineering and Computer Science, The University of Queensland, Saint Lucia, Brisbane, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2026,5,23]]},"reference":[{"key":"e_1_3_2_2_2","doi-asserted-by":"publisher","DOI":"10.1137\/S0036144598347497"},{"key":"e_1_3_2_3_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.is.2024.102427"},{"key":"e_1_3_2_4_2","unstructured":"Benjamin Paul Chamberlain Stephen R. Hardwick David R. Wardrope Fabon Dzogang Fabio Daolio and Sa\u00fal Vargas. 2019. Scalable hyperbolic recommender systems. arXiv:1902.08648. Retrieved from https:\/\/arxiv.org\/abs\/1902.08648"},{"key":"e_1_3_2_5_2","article-title":"Hyperbolic graph convolutional neural networks","volume":"32","author":"Chami Ines","year":"2019","unstructured":"Ines Chami, Zhitao Ying, Christopher R\u00e9, and Jure Leskovec. 2019. Hyperbolic graph convolutional neural networks. In Advances in Neural Information Processing Systems, Vol. 32.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_3_2_6_2","doi-asserted-by":"publisher","DOI":"10.1145\/3460231.3474255"},{"key":"e_1_3_2_7_2","doi-asserted-by":"publisher","DOI":"10.1145\/3459637.3482242"},{"key":"e_1_3_2_8_2","unstructured":"Evgeny Frolov Lina Bashaeva Leyla Mirvakhabova and Ivan Oseledets. 2024. Hyperbolic embeddings in sequential self-attention for improved next-item recommendations. Retrieved April 29 2026 from https:\/\/openreview.net\/forum?id=0TZs6WOs16"},{"key":"e_1_3_2_9_2","volume-title":"Advances in Neural Information Processing Systems","volume":"31","author":"Ganea Octavian","year":"2018","unstructured":"Octavian Ganea, Gary B\u00e9cigneul, and Thomas Hofmann. 2018. Hyperbolic neural networks. In Advances in Neural Information Processing Systems, Vol. 31. Retrieved from https:\/\/proceedings.neurips.cc\/paper\/2018\/file\/ffeed84c7cb1ae957bfc4d9000fa9f38-Paper.pdf"},{"key":"e_1_3_2_10_2","unstructured":"Albert Gu and Tri Dao. 2023. Mamba: Linear-time sequence modeling with selective state spaces. arXiv:2312.00752. Retrieved from https:\/\/arxiv.org\/abs\/2312.00752"},{"key":"e_1_3_2_11_2","doi-asserted-by":"publisher","DOI":"10.1145\/3568395"},{"key":"e_1_3_2_12_2","unstructured":"B. Hidasi. 2015. Session-based recommendations with recurrent neural networks. arXiv:1511.06939. Retrieved from https:\/\/arxiv.org\/abs\/1511.06939"},{"key":"e_1_3_2_13_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICDM.2018.00035"},{"key":"e_1_3_2_14_2","doi-asserted-by":"publisher","DOI":"10.1145\/3132847.3132926"},{"key":"e_1_3_2_15_2","unstructured":"Jiacheng Li Jingbo Shang and Julian McAuley. 2022. UCTopic: Unsupervised contrastive learning for phrase representations and topic mining. arXiv:2202.13469. Retrieved from https:\/\/arxiv.org\/abs\/2202.13469"},{"key":"e_1_3_2_16_2","doi-asserted-by":"publisher","DOI":"10.1145\/3459637.3482351"},{"key":"e_1_3_2_17_2","unstructured":"Chengkai Liu Jianghao Lin Jianling Wang Hanzhou Liu and James Caverlee. 2024. Mamba4Rec: Towards efficient sequential recommendation with selective state space models. arXiv:2403.03900. Retrieved from https:\/\/arxiv.org\/abs\/2403.03900"},{"issue":"2","key":"e_1_3_2_18_2","first-page":"64","article-title":"Recurrent neural networks","volume":"5","author":"Medsker Larry R.","year":"2001","unstructured":"Larry R. Medsker and Lakhmi Jain. 2001. Recurrent neural networks. Design and Applications 5, 2 (2001), 64\u201367.","journal-title":"Design and Applications"},{"key":"e_1_3_2_19_2","doi-asserted-by":"publisher","DOI":"10.1145\/3383313.3412219"},{"key":"e_1_3_2_20_2","first-page":"24925","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Mishne Gal","year":"2023","unstructured":"Gal Mishne, Zhengchao Wan, Yusu Wang, and Sheng Yang. 2023. The numerical stability of hyperbolic representation learning. In Proceedings of the International Conference on Machine Learning. PMLR, 24925\u201324949."},{"key":"e_1_3_2_21_2","volume-title":"Advances in Neural Information Processing Systems","volume":"30","author":"Nickel Maximillian","year":"2017","unstructured":"Maximillian Nickel and Douwe Kiela. 2017. Poincar\u00e9 embeddings for learning hierarchical representations. In Advances in Neural Information Processing Systems, Vol. 30."},{"key":"e_1_3_2_22_2","first-page":"3779","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Nickel Maximillian","year":"2018","unstructured":"Maximillian Nickel and Douwe Kiela. 2018. Learning continuous hierarchies in the Lorentz model of hyperbolic geometry. In Proceedings of the International Conference on Machine Learning. PMLR, 3779\u20133788."},{"key":"e_1_3_2_23_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3136921"},{"key":"e_1_3_2_24_2","doi-asserted-by":"publisher","DOI":"10.1145\/3190616"},{"key":"e_1_3_2_25_2","unstructured":"Steffen Rendle Christoph Freudenthaler Zeno Gantner and Lars Schmidt-Thieme. 2012. BPR: Bayesian personalized ranking from implicit feedback. arXiv:1205.2618. Retrieved from https:\/\/arxiv.org\/abs\/1205.2618"},{"key":"e_1_3_2_26_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.physd.2019.132306"},{"key":"e_1_3_2_27_2","doi-asserted-by":"publisher","DOI":"10.1145\/3357384.3357895"},{"key":"e_1_3_2_28_2","doi-asserted-by":"publisher","DOI":"10.1145\/3442381.3450101"},{"key":"e_1_3_2_29_2","doi-asserted-by":"publisher","DOI":"10.1145\/3159652.3159656"},{"key":"e_1_3_2_30_2","volume-title":"Advances in Neural Information Processing Systems","volume":"30","author":"Vaswani Ashish","year":"2017","unstructured":"Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, \u0141ukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in Neural Information Processing Systems, Vol. 30."},{"key":"e_1_3_2_31_2","doi-asserted-by":"publisher","DOI":"10.1145\/3336191.3371850"},{"key":"e_1_3_2_32_2","doi-asserted-by":"crossref","unstructured":"Shoujin Wang Liang Hu Yan Wang Longbing Cao Quan Z. Sheng and Mehmet Orgun. 2019. Sequential recommender systems: Challenges progress and prospects. arXiv:2001.04830. Retrieved from https:\/\/arxiv.org\/abs\/2001.04830","DOI":"10.24963\/ijcai.2019\/883"},{"key":"e_1_3_2_33_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.3301346"},{"key":"e_1_3_2_34_2","doi-asserted-by":"publisher","DOI":"10.1145\/3539618.3592036"},{"key":"e_1_3_2_35_2","unstructured":"Jiyuan Yang Yuanzi Li Jingyu Zhao Hanbing Wang Muyang Ma Jun Ma Zhaochun Ren Mengqi Zhang Xin Xin Zhumin Chen and Pengjie Ren.2024. Uncovering selective state space model\u2019s capabilities in lifelong sequential recommendation. arXiv:2403.16371. Retrieved from https:\/\/arxiv.org\/abs\/2403.16371"},{"key":"e_1_3_2_36_2","doi-asserted-by":"publisher","DOI":"10.1145\/3637528.3672039"},{"key":"e_1_3_2_37_2","doi-asserted-by":"publisher","DOI":"10.1145\/3626772.3657762"},{"key":"e_1_3_2_38_2","doi-asserted-by":"publisher","DOI":"10.1145\/2487575.2487608"},{"key":"e_1_3_2_39_2","doi-asserted-by":"publisher","DOI":"10.1145\/3616855.3635760"},{"key":"e_1_3_2_40_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICDE60146.2024.00049"}],"container-title":["ACM Transactions on Information Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3811405","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,23]],"date-time":"2026-05-23T12:29:58Z","timestamp":1779539398000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3811405"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,23]]},"references-count":39,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2026,6,30]]}},"alternative-id":["10.1145\/3811405"],"URL":"https:\/\/doi.org\/10.1145\/3811405","relation":{},"ISSN":["1046-8188","1558-2868"],"issn-type":[{"value":"1046-8188","type":"print"},{"value":"1558-2868","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,23]]},"assertion":[{"value":"2025-04-15","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-04-15","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-05-23","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}