{"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-contextual-attention-based-gru","title":"A Hierarchical Contextual Attention-based GRU Network for Sequential Recommendation","arxiv_id":"1711.05114","date":"2017-11-14","proceeding":null,"authors":["Qiang Cui","Shu Wu","Yan Huang","Liang Wang"],"abstract":"Sequential recommendation is one of fundamental tasks for Web applications.\nPrevious methods are mostly based on Markov chains with a strong Markov\nassumption. Recently, recurrent neural networks (RNNs) are getting more and\nmore popular and has demonstrated its effectiveness in many tasks. The last\nhidden state is usually applied as the sequence's representation to make\nrecommendation. Benefit from the natural characteristics of RNN, the hidden\nstate is a combination of long-term dependency and short-term interest to some\ndegrees. However, the monotonic temporal dependency of RNN impairs the user's\nshort-term interest. Consequently, the hidden state is not sufficient to\nreflect the user's final interest. In this work, to deal with this problem, we\npropose a Hierarchical Contextual Attention-based GRU (HCA-GRU) network. The\nfirst level of HCA-GRU is conducted on the input. We construct a contextual\ninput by using several recent inputs based on the attention mechanism. This can\nmodel the complicated correlations among recent items and strengthen the hidden\nstate. The second level is executed on the hidden state. We fuse the current\nhidden state and a contextual hidden state built by the attention mechanism,\nwhich leads to a more suitable user's overall interest. Experiments on two\nreal-world datasets show that HCA-GRU can effectively generate the personalized\nranking list and achieve significant improvement.","url_abs":"http://arxiv.org/abs/1711.05114v3","url_pdf":"http://arxiv.org/pdf/1711.05114v3.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-contextual-attention-based-gru","repo_url":"https://github.com/cuiqiang1990/HCA","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"sequential-recommendation","task_name":"Sequential Recommendation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}