{"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/neural-response-ranking-for-social","title":"Neural Response Ranking for Social Conversation: A Data-Efficient Approach","arxiv_id":"1811.00967","date":"2018-11-02","proceeding":"WS 2018 10","authors":["Igor Shalyminov","Ondřej Dušek","Oliver Lemon"],"abstract":"The overall objective of 'social' dialogue systems is to support engaging,\nentertaining, and lengthy conversations on a wide variety of topics, including\nsocial chit-chat. Apart from raw dialogue data, user-provided ratings are the\nmost common signal used to train such systems to produce engaging responses. In\nthis paper we show that social dialogue systems can be trained effectively from\nraw unannotated data. Using a dataset of real conversations collected in the\n2017 Alexa Prize challenge, we developed a neural ranker for selecting 'good'\nsystem responses to user utterances, i.e. responses which are likely to lead to\nlong and engaging conversations. We show that (1) our neural ranker\nconsistently outperforms several strong baselines when trained to optimise for\nuser ratings; (2) when trained on larger amounts of data and only using\nconversation length as the objective, the ranker performs better than the one\ntrained using ratings -- ultimately reaching a Precision@1 of 0.87. This\nadvance will make data collection for social conversational agents simpler and\nless expensive in the future.","url_abs":"http://arxiv.org/abs/1811.00967v1","url_pdf":"http://arxiv.org/pdf/1811.00967v1.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":"neural-response-ranking-for-social","repo_url":"https://github.com/WattSocialBot/alana_learning_to_rank","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.00967","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}