{"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/a-hierarchical-attention-model-for-social","title":"A Hierarchical Attention Model for Social Contextual Image Recommendation","arxiv_id":"1806.00723","date":"2018-06-03","proceeding":null,"authors":["Le Wu","Lei Chen","Richang Hong","Yanjie Fu","Xing Xie","Meng Wang"],"abstract":"Image based social networks are among the most popular social networking\nservices in recent years. With tremendous images uploaded everyday,\nunderstanding users' preferences on user-generated images and making\nrecommendations have become an urgent need. In fact, many hybrid models have\nbeen proposed to fuse various kinds of side information~(e.g., image visual\nrepresentation, social network) and user-item historical behavior for enhancing\nrecommendation performance. However, due to the unique characteristics of the\nuser generated images in social image platforms, the previous studies failed to\ncapture the complex aspects that influence users' preferences in a unified\nframework. Moreover, most of these hybrid models relied on predefined weights\nin combining different kinds of information, which usually resulted in\nsub-optimal recommendation performance. To this end, in this paper, we develop\na hierarchical attention model for social contextual image recommendation. In\naddition to basic latent user interest modeling in the popular matrix\nfactorization based recommendation, we identify three key aspects (i.e., upload\nhistory, social influence, and owner admiration) that affect each user's latent\npreferences, where each aspect summarizes a contextual factor from the complex\nrelationships between users and images. After that, we design a hierarchical\nattention network that naturally mirrors the hierarchical relationship\n(elements in each aspects level, and the aspect level) of users' latent\ninterests with the identified key aspects. Specifically, by taking embeddings\nfrom state-of-the-art deep learning models that are tailored for each kind of\ndata, the hierarchical attention network could learn to attend differently to\nmore or less content. Finally, extensive experimental results on real-world\ndatasets clearly show the superiority of our proposed model.","url_abs":"http://arxiv.org/abs/1806.00723v3","url_pdf":"http://arxiv.org/pdf/1806.00723v3.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":"a-hierarchical-attention-model-for-social","repo_url":"https://github.com/newlei/HASC","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}