{"id":627885,"date":"2019-12-18T18:05:26","date_gmt":"2019-12-19T02:05:26","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-research-item&#038;p=627885"},"modified":"2019-12-18T18:05:26","modified_gmt":"2019-12-19T02:05:26","slug":"data-anonymous-encoding-for-text-to-sql-generation","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/data-anonymous-encoding-for-text-to-sql-generation\/","title":{"rendered":"Data-Anonymous Encoding for Text-to-SQL Generation"},"content":{"rendered":"<p><span>On text-to-SQL generation, the input utterance usually contains lots of tokens that are related to column names or cells in the table, called table-related tokens. These table-related tokens are troublesome for the downstream neural semantic parser because it brings complex semantics and hinders the sharing across the training examples. However, existing approaches either ignore handling these tokens before the semantic parser or simply use deterministic approaches based on string-match or word embedding similarity. In this work, we propose a more efficient approach to handle table-related tokens before the semantic parser. First, we formulate it as a sequential tagging problem and propose a two-stage anonymization model to learn the semantic relationship between tables and input utterances. Then, we leverage the implicit supervision from SQL queries by policy gradient to guide the training. Experiments demonstrate that our approach consistently improves performances of different neural semantic parsers and significantly outperforms deterministic approaches.<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>On text-to-SQL generation, the input utterance usually contains lots of tokens that are related to column names or cells in the table, called table-related tokens. These table-related tokens are troublesome for the downstream neural semantic parser because it brings complex semantics and hinders the sharing across the training examples. However, existing approaches either ignore handling [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","footnotes":""},"msr-content-type":[3],"msr-research-highlight":[],"research-area":[13556],"msr-publication-type":[193716],"msr-product-type":[],"msr-focus-area":[],"msr-platform":[],"msr-download-source":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-627885","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-artificial-intelligence","msr-locale-en_us"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2019-11","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":"ACL","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":"file","viewUrl":"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/12\/EMNLP2019_DataBinding_CameraReady.pdf","id":"627888","title":"emnlp2019_databinding_cameraready","label_id":"243109","label":0}],"msr_related_uploader":"","msr_attachments":[{"id":627888,"url":"https:\/\/www.microsoft.com\/en-us\/research\/uploads\/prod\/2019\/12\/EMNLP2019_DataBinding_CameraReady.pdf"}],"msr-author-ordering":[{"type":"text","value":"Zhen DONG","user_id":0,"rest_url":false},{"type":"text","value":"Shizhao SUN","user_id":0,"rest_url":false},{"type":"text","value":"Hongzhi LIU","user_id":0,"rest_url":false},{"type":"user_nicename","value":"Jian-Guang Lou","user_id":32337,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Jian-Guang Lou"},{"type":"user_nicename","value":"Dongmei Zhang","user_id":31665,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Dongmei Zhang"}],"msr_impact_theme":[],"msr_research_lab":[199560],"msr_event":[],"msr_group":[],"msr_project":[578947],"publication":[],"video":[],"download":[],"msr_publication_type":"inproceedings","related_content":{"projects":[{"ID":578947,"post_title":"Natural Language Interface for Data Analytics","post_name":"conversational-data-analytics","post_type":"msr-project","post_date":"2019-04-15 15:23:36","post_modified":"2022-03-22 02:54:11","post_status":"publish","permalink":"https:\/\/www.microsoft.com\/en-us\/research\/project\/conversational-data-analytics\/","post_excerpt":"In this project, we try to research and develop a conversation technology for data analytics scenarios. By using our technology, given a relational database or a data table, a user can explore the data table and insights from the dataset through natural language conversation. Our system can understand user\u2019s natural language questions and convert the questions into some analysis programs. The programs can be executed on the relational database (or the data table) to obtain&hellip;","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/578947"}]}}]},"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/627885"}],"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":1,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/627885\/revisions"}],"predecessor-version":[{"id":627891,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/627885\/revisions\/627891"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=627885"}],"wp:term":[{"taxonomy":"msr-content-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-content-type?post=627885"},{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=627885"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=627885"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=627885"},{"taxonomy":"msr-product-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-product-type?post=627885"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=627885"},{"taxonomy":"msr-platform","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-platform?post=627885"},{"taxonomy":"msr-download-source","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-download-source?post=627885"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=627885"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=627885"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=627885"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=627885"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=627885"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=627885"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=627885"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}