{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deep-bayesian-multi-target-learning-for","title":"Deep Bayesian Multi-Target Learning for Recommender Systems","arxiv_id":"1902.09154","date":"2019-02-25","proceeding":null,"authors":["Qi. Wang","Zhihui Ji","Huasheng Liu","Binqiang Zhao"],"abstract":"With the increasing variety of services that e-commerce platforms provide,\ncriteria for evaluating their success become also increasingly multi-targeting.\nThis work introduces a multi-target optimization framework with Bayesian\nmodeling of the target events, called Deep Bayesian Multi-Target Learning\n(DBMTL). In this framework, target events are modeled as forming a Bayesian\nnetwork, in which directed links are parameterized by hidden layers, and\nlearned from training samples. The structure of Bayesian network is determined\nby model selection. We applied the framework to Taobao live-streaming\nrecommendation, to simultaneously optimize (and strike a balance) on targets\nincluding click-through rate, user stay time in live room, purchasing behaviors\nand interactions. Significant improvement has been observed for the proposed\nmethod over other MTL frameworks and the non-MTL model. Our practice shows that\nwith an integrated causality structure, we can effectively make the learning of\na target benefit from other targets, creating significant synergy effects that\nimprove all targets. The neural network construction guided by DBMTL fits in\nwith the general probabilistic model connecting features and multiple targets,\ntaking weaker assumption than the other methods discussed in this paper. This\ntheoretical generality brings about practical generalization power over various\ntargets distributions, including sparse targets and continuous-value ones.","url_abs":"http://arxiv.org/abs/1902.09154v1","url_pdf":"http://arxiv.org/pdf/1902.09154v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deep-bayesian-multi-target-learning-for","repo_url":"https://github.com/alibaba/EasyRec/blob/master/easy_rec/python/model/dbmtl.py","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null},{"paper_slug":"deep-bayesian-multi-target-learning-for","repo_url":"https://github.com/alibaba/TorchEasyRec","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"model-selection","task_name":"Model Selection"},{"task_slug":"recommendation-systems","task_name":"Recommendation Systems"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}