{"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/dkn-deep-knowledge-aware-network-for-news","title":"DKN: Deep Knowledge-Aware Network for News Recommendation","arxiv_id":"1801.08284","date":"2018-01-25","proceeding":null,"authors":["Hongwei Wang","Fuzheng Zhang","Xing Xie","Minyi Guo"],"abstract":"Online news recommender systems aim to address the information explosion of\nnews and make personalized recommendation for users. In general, news language\nis highly condensed, full of knowledge entities and common sense. However,\nexisting methods are unaware of such external knowledge and cannot fully\ndiscover latent knowledge-level connections among news. The recommended results\nfor a user are consequently limited to simple patterns and cannot be extended\nreasonably. Moreover, news recommendation also faces the challenges of high\ntime-sensitivity of news and dynamic diversity of users' interests. To solve\nthe above problems, in this paper, we propose a deep knowledge-aware network\n(DKN) that incorporates knowledge graph representation into news\nrecommendation. DKN is a content-based deep recommendation framework for\nclick-through rate prediction. The key component of DKN is a multi-channel and\nword-entity-aligned knowledge-aware convolutional neural network (KCNN) that\nfuses semantic-level and knowledge-level representations of news. KCNN treats\nwords and entities as multiple channels, and explicitly keeps their alignment\nrelationship during convolution. In addition, to address users' diverse\ninterests, we also design an attention module in DKN to dynamically aggregate a\nuser's history with respect to current candidate news. Through extensive\nexperiments on a real online news platform, we demonstrate that DKN achieves\nsubstantial gains over state-of-the-art deep recommendation models. We also\nvalidate the efficacy of the usage of knowledge in DKN.","url_abs":"http://arxiv.org/abs/1801.08284v2","url_pdf":"http://arxiv.org/pdf/1801.08284v2.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":"dkn-deep-knowledge-aware-network-for-news","repo_url":"https://github.com/810410738/DKN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"dkn-deep-knowledge-aware-network-for-news","repo_url":"https://github.com/hwwang55/DKN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"dkn-deep-knowledge-aware-network-for-news","repo_url":"https://github.com/microsoft/recommenders","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null},{"paper_slug":"dkn-deep-knowledge-aware-network-for-news","repo_url":"https://github.com/yusanshi/NewsRecommendation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"click-through-rate-prediction","task_name":"Click-Through Rate Prediction"},{"task_slug":"common-sense-reasoning","task_name":"Common Sense Reasoning"},{"task_slug":"news-recommendation","task_name":"News Recommendation"},{"task_slug":"recommendation-systems","task_name":"Recommendation Systems"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/click-through-rate-prediction-on-bing-news","task":"Click-Through Rate Prediction","dataset":"Bing News","model":"DKN","rank_in_archive_order":6,"of":7,"metrics":{"AUC":"0.659"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1801.08284","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}