{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,12,24]],"date-time":"2025-12-24T12:09:39Z","timestamp":1766578179369,"version":"3.48.0"},"reference-count":52,"publisher":"Oxford University Press (OUP)","issue":"12","license":[{"start":{"date-parts":[[2025,6,23]],"date-time":"2025-06-23T00:00:00Z","timestamp":1750636800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/academic.oup.com\/pages\/standard-publication-reuse-rights"}],"funder":[{"DOI":"10.13039\/501100017691","name":"Key Research and Development Program of Guangxi","doi-asserted-by":"publisher","award":["GuikeAB22080065"],"award-info":[{"award-number":["GuikeAB22080065"]}],"id":[{"id":"10.13039\/501100017691","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation","doi-asserted-by":"crossref","award":["61762029"],"award-info":[{"award-number":["61762029"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,12,24]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Recently, click-through rate (CTR) prediction research based on empirical risk minimization has achieved remarkable results, which assumes that training and test data follow the same distribution and optimizes CTR prediction models by minimizing prediction errors in the training data. However, this assumption does not hold in the real world. Specifically, user preferences change over time, leading to \u2018temporal drift,\u2019 where the distribution of test data differs from that of the training data. In this paper, we propose a \u2018temporal drift adaptation framework\u2019 (TDAF) for CTR prediction to cope with temporal drift. In TDAF, we devise a feature embedding predictor to learn the evolution of user preferences from historical feature embeddings and simulate feature embeddings after temporal drift. The model\u2019s performance on simulated and real feature embeddings is improved through a novel bi-level optimization based on meta-learning, enhancing its ability to cope with temporal drift. TDAF is model-agnostic and can be widely applied to common CTR prediction models. We conduct extensive experiments, and the results show that TDAF improves the performance of CTR prediction models by an average of 0.55% across multiple datasets. Theoretical analysis and ablation studies further validate the effectiveness of TDAF. The code is available at https:\/\/github.com\/lgxccc\/TDAF-for-CTR-prediction\/tree\/main.<\/jats:p>","DOI":"10.1093\/comjnl\/bxaf079","type":"journal-article","created":{"date-parts":[[2025,6,6]],"date-time":"2025-06-06T07:51:41Z","timestamp":1749196301000},"page":"1857-1869","source":"Crossref","is-referenced-by-count":0,"title":["TDAF: a bi-level optimization framework for CTR prediction with temporal drift adaptation"],"prefix":"10.1093","volume":"68","author":[{"given":"Guoxin","family":"Lu","sequence":"first","affiliation":[{"name":"School of Business, Guilin University of Electronic Technology , P.C. 541004, No. 1 Jinji Road, Qixing District, Guilin ,","place":["China"]}]},{"given":"Pingshan","family":"Liu","sequence":"additional","affiliation":[{"name":"School of Business, Guilin University of Electronic Technology , P.C. 541004, No. 1 Jinji Road, Qixing District, Guilin ,","place":["China"]}]},{"given":"Zhen","family":"Yang","sequence":"additional","affiliation":[{"name":"School of Business, Guilin University of Electronic Technology , P.C. 541004, No. 1 Jinji Road, Qixing District, Guilin ,","place":["China"]}]}],"member":"286","published-online":{"date-parts":[[2025,6,23]]},"reference":[{"key":"2025122407071574700_ref1","doi-asserted-by":"crossref","first-page":"805","DOI":"10.1145\/3539597.3570372","article-title":"Cl4ctr: a contrastive learning framework for CTR prediction","volume-title":"Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining","author":"Wang","year":"2023"},{"key":"2025122407071574700_ref2","doi-asserted-by":"crossref","first-page":"846","DOI":"10.1145\/3616855.3635829","article-title":"Deep evolutional instant interest network for CTR prediction in trigger-induced recommendation","volume-title":"Proceedings of the 17th ACM International Conference on Web Search and Data Mining","author":"Xiao","year":"2024"},{"volume-title":"Statistical Learning Theory","year":"1998","author":"Vapnik","key":"2025122407071574700_ref3"},{"key":"2025122407071574700_ref4","first-page":"1725","article-title":"Deepfm: a factorization-machine based neural network for CTR prediction","volume-title":"Proceedings of the 26th International Joint Conference on Artificial Intelligence","author":"Guo","year":"2017"},{"key":"2025122407071574700_ref5","first-page":"7","article-title":"Deep & cross network for ad click predictions","volume-title":"Proceedings of the ADKDD\u201917","author":"Wang","year":"2017"},{"key":"2025122407071574700_ref6","doi-asserted-by":"crossref","first-page":"1059","DOI":"10.1145\/3219819.3219823","article-title":"Deep interest network for click-through rate prediction","volume-title":"Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining","author":"Zhou","year":"2018"},{"key":"2025122407071574700_ref7","doi-asserted-by":"crossref","first-page":"1386","DOI":"10.1145\/3539618.3591755","article-title":"Reformulating ctr prediction: learning invariant feature interactions for recommendation","volume-title":"Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval","author":"Zhang","year":"2023"},{"key":"2025122407071574700_ref8","doi-asserted-by":"crossref","first-page":"331","DOI":"10.1145\/3539618.3591624","article-title":"A generic learning framework for sequential recommendation with distribution shifts","volume-title":"Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval","author":"Yang","year":"2023"},{"key":"2025122407071574700_ref9","first-page":"428","article-title":"Generalize for future: slow and fast trajectory learning for CTR prediction","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Zhu","year":"2024"},{"key":"2025122407071574700_ref10","first-page":"1149","article-title":"Product-based neural networks for user response prediction","volume-title":"2016 IEEE 16th International Conference on Data Mining (ICDM), Los Alamitos, CA, United States","author":"Qu","year":"2016"},{"key":"2025122407071574700_ref11","first-page":"3119","article-title":"Attentional factorization machines: learning the weight of feature interactions via attention networks","volume-title":"Proceedings of the 26th International Joint Conference on Artificial Intelligence","author":"Xiao","year":"2017"},{"key":"2025122407071574700_ref12","doi-asserted-by":"publisher","first-page":"2913","DOI":"10.1007\/s11042-023-15765-4","article-title":"A domain generalization pedestrian re-identification algorithm based on meta-graph aware","volume":"83","author":"Wu","year":"2024","journal-title":"Multimed Tools Appl"},{"key":"2025122407071574700_ref13","first-page":"11147","article-title":"Foresee what you will learn: data augmentation for domain generalization in non-stationary environment","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Zeng","year":"2023"},{"key":"2025122407071574700_ref14","article-title":"Latent trajectory learning for limited timestamps under distribution shift over time","volume-title":"The Twelfth International Conference on Learning Representations","author":"Zeng","year":"2024"},{"key":"2025122407071574700_ref15","first-page":"1126","article-title":"Model-agnostic meta-learning for fast adaptation of deep networks","volume-title":"International Conference on Machine Learning","author":"Finn","year":"2017"},{"key":"2025122407071574700_ref16","doi-asserted-by":"crossref","DOI":"10.1609\/aaai.v32i1.11596","article-title":"Learning to generalize: meta-learning for domain generalization","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Li","year":"2018"},{"key":"2025122407071574700_ref17","doi-asserted-by":"crossref","first-page":"521","DOI":"10.1145\/1242572.1242643","article-title":"Predicting clicks: estimating the click-through rate for new ads","volume-title":"Proceedings of the 16th International Conference on World Wide Web","author":"Richardson","year":"2007"},{"key":"2025122407071574700_ref18","doi-asserted-by":"crossref","first-page":"45","DOI":"10.1007\/978-3-319-30671-1_4","article-title":"Deep learning over multi-field categorical data: \u2013a case study on user response prediction","volume-title":"Advances in Information Retrieval: 38th European Conference on IR Research","author":"Zhang","year":"2016"},{"key":"2025122407071574700_ref19"},{"key":"2025122407071574700_ref20","doi-asserted-by":"crossref","first-page":"995","DOI":"10.1109\/ICDM.2010.127","article-title":"Factorization machines","volume-title":"2010 IEEE International Conference on Data Mining, Los Alamitos, CA, United States","author":"Rendle","year":"2010"},{"key":"2025122407071574700_ref21"},{"key":"2025122407071574700_ref22","doi-asserted-by":"crossref","first-page":"1754","DOI":"10.1145\/3219819.3220023","article-title":"Xdeepfm: combining explicit and implicit feature interactions for recommender systems","volume-title":"Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining","author":"Lian","year":"2018"},{"key":"2025122407071574700_ref23","first-page":"4552","article-title":"Finalmlp: an enhanced two-stream mlp model for ctr prediction","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Mao","year":"2023"},{"key":"2025122407071574700_ref24","doi-asserted-by":"crossref","DOI":"10.1609\/aaai.v28i1.8917","article-title":"Sequential click prediction for sponsored search with recurrent neural networks","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Zhang","year":"2014"},{"key":"2025122407071574700_ref25","doi-asserted-by":"crossref","first-page":"1743","DOI":"10.1145\/2806416.2806603","article-title":"A convolutional click prediction model","volume-title":"Proceedings of the 24th ACM International on Conference on Information and Knowledge Management","author":"Liu","year":"2015"},{"key":"2025122407071574700_ref26","first-page":"5941","article-title":"Deep interest evolution network for click-through rate prediction","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Zhou","year":"2019"},{"key":"2025122407071574700_ref27","article-title":"Neural machine translation by jointly learning to align and translate","volume-title":"3rd International Conference on Learning Representations, ICLR 2015","author":"Bahdanau","year":"2015"},{"key":"2025122407071574700_ref28"},{"key":"2025122407071574700_ref29"},{"key":"2025122407071574700_ref30"},{"key":"2025122407071574700_ref31","first-page":"21983","article-title":"Invariant learning via probability of sufficient and necessary causes","volume":"36","author":"Yang","year":"2024","journal-title":"Adv Neural Inform Process Syst"},{"key":"2025122407071574700_ref32","first-page":"6804","article-title":"Heterogeneous risk minimization","volume-title":"International Conference on Machine Learning","author":"Liu","year":"2021"},{"key":"2025122407071574700_ref33","first-page":"2189","article-title":"Environment inference for invariant learning","volume-title":"International Conference on Machine Learning","author":"Creager","year":"2021"},{"key":"2025122407071574700_ref34"},{"key":"2025122407071574700_ref35","doi-asserted-by":"publisher","first-page":"1378","DOI":"10.1214\/20-AOS2004","article-title":"Learning models with uniform performance via distributionally robust optimization","volume":"49","author":"Duchi","year":"2021","journal-title":"Ann Stat"},{"key":"2025122407071574700_ref36","doi-asserted-by":"publisher","first-page":"603","DOI":"10.1287\/moor.2022.1275","article-title":"Distributionally robust stochastic optimization with wasserstein distance","volume":"48","author":"Gao","year":"2023","journal-title":"Math Oper Res"},{"key":"2025122407071574700_ref37","doi-asserted-by":"publisher","first-page":"5362","DOI":"10.1109\/TPAMI.2024.3367329","article-title":"A comprehensive survey of continual learning: theory, method and application","volume":"46","author":"Wang","year":"2024","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"2025122407071574700_ref38","first-page":"21983","article-title":"Evolving standardization for continual domain generalization over temporal drift","volume":"36","author":"Xie","year":"2024","journal-title":"Adv Neural Inform Process Syst"},{"key":"2025122407071574700_ref39","first-page":"12773","article-title":"Instance-conditional timescales of decay for non-stationary learning","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Jain","year":"2024"},{"key":"2025122407071574700_ref40","doi-asserted-by":"crossref","DOI":"10.1145\/3690624.3709299","article-title":"Generalizable recommender system during temporal popularity distribution shifts","volume-title":"Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining","author":"Yoo","year":"2025"},{"key":"2025122407071574700_ref41","first-page":"5998","article-title":"Attention is all you need","volume":"30","author":"Vaswani","year":"2017","journal-title":"Adv Neural Inform Process Syst"},{"key":"2025122407071574700_ref42","first-page":"20029","article-title":"PMR: prototypical modal rebalance for multimodal learning","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Fan","year":"2023"},{"key":"2025122407071574700_ref43","first-page":"16096","article-title":"Domain generalization via entropy regularization","volume":"33","author":"Zhao","year":"2020","journal-title":"Adv Neural Inform Process Syst"},{"key":"2025122407071574700_ref44"},{"key":"2025122407071574700_ref45","doi-asserted-by":"publisher","first-page":"2685","DOI":"10.1007\/s10994-023-06324-x","article-title":"Domain adversarial neural networks for domain generalization: when it works and how to improve","volume":"112","author":"Sicilia","year":"2023","journal-title":"Mach Learn"},{"key":"2025122407071574700_ref46","doi-asserted-by":"publisher","first-page":"151","DOI":"10.1007\/s10994-009-5152-4","article-title":"A theory of learning from different domains","volume":"79","author":"Ben-David","year":"2010","journal-title":"Mach Learn"},{"key":"2025122407071574700_ref47","first-page":"1","article-title":"The movielens datasets: history and context","volume":"5","author":"Harper","year":"2015","journal-title":"ACM Trans Interact Intell. Syst. (TIIS)"},{"key":"2025122407071574700_ref48","doi-asserted-by":"crossref","first-page":"3562","DOI":"10.1145\/3485447.3512251","article-title":"Causal representation learning for out-of-distribution recommendation","volume-title":"Proceedings of the ACM Web Conference 2022","author":"Wang","year":"2022"},{"key":"2025122407071574700_ref49"},{"volume-title":"The Twelfth International Conference on Learning Representations","key":"2025122407071574700_ref50"},{"key":"2025122407071574700_ref51","first-page":"10309","article-title":"Wild-time: a benchmark of in-the-wild distribution shift over time","volume":"35","author":"Yao","year":"2022","journal-title":"Adv Neural Inform Process Syst"},{"key":"2025122407071574700_ref52","doi-asserted-by":"crossref","first-page":"3173","DOI":"10.1145\/3589334.3645351","article-title":"Rethinking cross-domain sequential recommendation under open-world assumptions","volume-title":"Proceedings of the ACM on Web Conference 2024","author":"Xu","year":"2024"}],"container-title":["The Computer Journal"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/comjnl\/article-pdf\/68\/12\/1857\/63547638\/bxaf079.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/comjnl\/article-pdf\/68\/12\/1857\/63547638\/bxaf079.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,24]],"date-time":"2025-12-24T12:07:42Z","timestamp":1766578062000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/comjnl\/article\/68\/12\/1857\/8171718"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,23]]},"references-count":52,"journal-issue":{"issue":"12","published-online":{"date-parts":[[2025,6,23]]},"published-print":{"date-parts":[[2025,12,24]]}},"URL":"https:\/\/doi.org\/10.1093\/comjnl\/bxaf079","relation":{},"ISSN":["0010-4620","1460-2067"],"issn-type":[{"type":"print","value":"0010-4620"},{"type":"electronic","value":"1460-2067"}],"subject":[],"published-other":{"date-parts":[[2025,12]]},"published":{"date-parts":[[2025,6,23]]}}}