While neural conversation models have shown great potentials towards generating
\ninformative and engaging responses via introducing external knowledge, learning
\nsuch a model often requires knowledge-grounded dialogues that are difficult to
\nobtain. To overcome the data challenge and reduce the cost of building a knowledge grounded dialogue system, we explore the problem under a zero-resource setting
\nby assuming no context-knowledge-response triples are needed for training. To
\nthis end, we propose representing the knowledge that bridges a context and a
\nresponse and the way that the knowledge is expressed as latent variables, and
\ndevise a variational approach that can effectively estimate a generation model
\nfrom a dialogue corpus and a knowledge corpus that are independent with each
\nother. Evaluation results on three benchmarks of knowledge-grounded dialogue
\ngeneration indicate that our model can achieve comparable performance with state of-the-art methods that rely on knowledge-grounded dialogues for training, and
\nexhibits a good generalization ability over different topics and different datasets.
\nCode is available at https:\/\/github.com\/nlpxucan\/ZRKGC.<\/p>\n","protected":false},"excerpt":{"rendered":"
While neural conversation models have shown great potentials towards generating informative and engaging responses via introducing external knowledge, learning such a model often requires knowledge-grounded dialogues that are difficult to obtain. To overcome the data challenge and reduce the cost of building a knowledge grounded dialogue system, we explore the problem under a zero-resource setting […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"footnotes":""},"msr-content-type":[3],"msr-research-highlight":[],"research-area":[13556,13545],"msr-publication-type":[193716],"msr-product-type":[],"msr-focus-area":[],"msr-platform":[],"msr-download-source":[],"msr-locale":[268875],"msr-field-of-study":[246691,247813],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-727627","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-artificial-intelligence","msr-research-area-human-language-technologies","msr-locale-en_us","msr-field-of-study-computer-science","msr-field-of-study-knowledge-management"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2020-7-1","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"url","viewUrl":"false","id":"false","title":"http:\/\/proceedings.neurips.cc\/paper\/2020\/file\/609c5e5089a9aa967232aba2a4d03114-Paper.pdf","label_id":"243132","label":0},{"type":"url","viewUrl":"false","id":"false","title":"https:\/\/papers.nips.cc\/paper\/2020\/file\/609c5e5089a9aa967232aba2a4d03114-Paper.pdf","label_id":"243132","label":0},{"type":"url","viewUrl":"false","id":"false","title":"https:\/\/dblp.uni-trier.de\/db\/conf\/nips\/neurips2020.html#LiX0ZZT20","label_id":"243109","label":0},{"type":"url","viewUrl":"false","id":"false","title":"https:\/\/papers.nips.cc\/paper\/2020\/hash\/609c5e5089a9aa967232aba2a4d03114-Abstract.html","label_id":"243109","label":0},{"type":"url","viewUrl":"false","id":"false","title":"https:\/\/proceedings.neurips.cc\/paper\/2020\/hash\/609c5e5089a9aa967232aba2a4d03114-Abstract.html","label_id":"243109","label":0}],"msr_related_uploader":"","msr_attachments":[],"msr-author-ordering":[{"type":"text","value":"Linxiao Li","user_id":0,"rest_url":false},{"type":"text","value":"Can Xu","user_id":0,"rest_url":false},{"type":"text","value":"Wei Wu","user_id":0,"rest_url":false},{"type":"text","value":"Yufan Zhao","user_id":0,"rest_url":false},{"type":"text","value":"Xueliang Zhao","user_id":0,"rest_url":false},{"type":"text","value":"Chongyang Tao","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[835921],"msr_project":[],"publication":[],"video":[],"download":[],"msr_publication_type":"inproceedings","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/727627"}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":2,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/727627\/revisions"}],"predecessor-version":[{"id":727648,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/727627\/revisions\/727648"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=727627"}],"wp:term":[{"taxonomy":"msr-content-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-content-type?post=727627"},{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=727627"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=727627"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=727627"},{"taxonomy":"msr-product-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-product-type?post=727627"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=727627"},{"taxonomy":"msr-platform","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-platform?post=727627"},{"taxonomy":"msr-download-source","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-download-source?post=727627"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=727627"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=727627"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=727627"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=727627"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=727627"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=727627"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}