{"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/better-conversations-by-modelingfilteringand","title":"Better Conversations by Modeling,Filtering,and Optimizing for Coherence and Diversity","arxiv_id":"1809.06873","date":"2018-09-18","proceeding":null,"authors":["Xinnuo Xu","Ondřej Dušek","Ioannis Konstas","Verena Rieser"],"abstract":"We present three enhancements to existing encoder-decoder models for\nopen-domain conversational agents, aimed at effectively modeling coherence and\npromoting output diversity: (1) We introduce a measure of coherence as the\nGloVe embedding similarity between the dialogue context and the generated\nresponse, (2) we filter our training corpora based on the measure of coherence\nto obtain topically coherent and lexically diverse context-response pairs, (3)\nwe then train a response generator using a conditional variational autoencoder\nmodel that incorporates the measure of coherence as a latent variable and uses\na context gate to guarantee topical consistency with the context and promote\nlexical diversity. Experiments on the OpenSubtitles corpus show a substantial\nimprovement over competitive neural models in terms of BLEU score as well as\nmetrics of coherence and diversity.","url_abs":"http://arxiv.org/abs/1809.06873v1","url_pdf":"http://arxiv.org/pdf/1809.06873v1.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":"better-conversations-by-modelingfilteringand","repo_url":"https://github.com/XinnuoXu/CVAE_Dial","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null},{"paper_slug":"better-conversations-by-modelingfilteringand","repo_url":"https://github.com/ricsinaruto/dialog-eval","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"diversity","task_name":"Diversity"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1809.06873","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}