{"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/product-based-neural-networks-for-user-1","title":"Product-based Neural Networks for User Response Prediction over Multi-field Categorical Data","arxiv_id":"1807.00311","date":"2018-07-01","proceeding":null,"authors":["Yanru Qu","Bohui Fang","Wei-Nan Zhang","Ruiming Tang","Minzhe Niu","Huifeng Guo","Yong Yu","Xiuqiang He"],"abstract":"User response prediction is a crucial component for personalized information\nretrieval and filtering scenarios, such as recommender system and web search.\nThe data in user response prediction is mostly in a multi-field categorical\nformat and transformed into sparse representations via one-hot encoding. Due to\nthe sparsity problems in representation and optimization, most research focuses\non feature engineering and shallow modeling. Recently, deep neural networks\nhave attracted research attention on such a problem for their high capacity and\nend-to-end training scheme. In this paper, we study user response prediction in\nthe scenario of click prediction. We first analyze a coupled gradient issue in\nlatent vector-based models and propose kernel product to learn field-aware\nfeature interactions. Then we discuss an insensitive gradient issue in\nDNN-based models and propose Product-based Neural Network (PNN) which adopts a\nfeature extractor to explore feature interactions. Generalizing the kernel\nproduct to a net-in-net architecture, we further propose Product-network In\nNetwork (PIN) which can generalize previous models. Extensive experiments on 4\nindustrial datasets and 1 contest dataset demonstrate that our models\nconsistently outperform 8 baselines on both AUC and log loss. Besides, PIN\nmakes great CTR improvement (relatively 34.67%) in online A/B test.","url_abs":"http://arxiv.org/abs/1807.00311v1","url_pdf":"http://arxiv.org/pdf/1807.00311v1.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":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/Atomu2014/product-nets-distributed","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/Atomu2014/product-nets","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/UlionTse/mlgb","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/jccarles/product-nets","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/DataCanvasIO/DeepTables","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/MindSpore-scientific-2/code-4/tree/main/PR_Product","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/MindSpore-scientific/code-13/tree/main/PR_Product","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/shenweichen/DeepCTR","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/shenweichen/DeepCTR-PyTorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"product-based-neural-networks-for-user-1","repo_url":"https://github.com/xue-pai/FuxiCTR","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":"feature-engineering","task_name":"Feature Engineering"},{"task_slug":"information-retrieval","task_name":"Information Retrieval"},{"task_slug":"prediction","task_name":"Prediction"},{"task_slug":"recommendation-systems","task_name":"Recommendation Systems"},{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1807.00311","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}