{"id":397838,"date":"2017-07-07T13:45:04","date_gmt":"2017-07-07T20:45:04","guid":{"rendered":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/?post_type=msr-research-item&#038;p=397838"},"modified":"2018-10-16T20:01:16","modified_gmt":"2018-10-17T03:01:16","slug":"joint-embedding-query-ad-leveraging-implicit-feedback","status":"publish","type":"msr-research-item","link":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/publication\/joint-embedding-query-ad-leveraging-implicit-feedback\/","title":{"rendered":"Joint Embedding of Query and Ad by Leveraging Implicit Feedback"},"content":{"rendered":"<p>Sponsored search is at the center of a multibillion\u00a0dollar market established by search technology.\u00a0Accurate ad click prediction is a key\u00a0component for this market to function since\u00a0the pricing mechanism heavily relies on the\u00a0estimation of click probabilities. Lexical features\u00a0derived from the text of both the query\u00a0and ads play a significant role, complementing\u00a0features based on historical click information.\u00a0The purpose of this paper is to explore the use\u00a0of word embedding techniques to generate effective\u00a0text features that can capture not only\u00a0lexical similarity between query and ads but\u00a0also the latent user intents. We identify several\u00a0potential weaknesses of the plain application\u00a0of conventional word embedding methodologies\u00a0for ad click prediction. These observations\u00a0motivated us to propose a set of novel\u00a0joint word embedding methods by leveraging\u00a0implicit click feedback. We verify the effectiveness\u00a0of these new word embedding models<br \/>\nby adding features derived from the new models\u00a0to the click prediction system of a commercial\u00a0search engine. Our evaluation results\u00a0clearly demonstrate the effectiveness of the\u00a0proposed methods. To the best of our knowledge\u00a0this work is the first successful application\u00a0of word embedding techniques for the task\u00a0of click prediction in sponsored search.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Sponsored search is at the center of a multibillion\u00a0dollar market established by search technology.\u00a0Accurate ad click prediction is a key\u00a0component for this market to function since\u00a0the pricing mechanism heavily relies on the\u00a0estimation of click probabilities. Lexical features\u00a0derived from the text of both the query\u00a0and ads play a significant role, complementing\u00a0features based on historical click information.\u00a0The [&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":"","msr-author-ordering":null,"msr_publishername":"Association for Computational Linguistics.","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"2015 Conference on Empirical Methods in Natural Language Processing, Lisbon, Portugal","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"","msr_journal":"","msr_number":"","msr_organization":"","msr_pages_string":"482\u2013491","msr_page_range_start":"482","msr_page_range_end":"491","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"2015 Conference on Empirical Methods in Natural Language Processing, Lisbon, Portugal","msr_doi":"","msr_arxiv_id":"","msr_s2_paper_id":"","msr_mag_id":"","msr_pubmed_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_original_fields_of_study":"","msr_release_tracker_id":"","msr_s2_match_type":"","msr_citation_count_updated":"","msr_published_date":"2015-09-17","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_s2_pdf_url":"","msr_year":0,"msr_citation_count":0,"msr_influential_citations":0,"msr_reference_count":0,"msr_s2_match_confidence":0,"msr_microsoftintellectualproperty":true,"msr_s2_open_access":false,"msr_s2_author_ids":[],"msr_pub_ids":[],"msr_hide_image_in_river":0,"footnotes":""},"msr-research-highlight":[],"research-area":[13555],"msr-publication-type":[193716],"msr-publisher":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-397838","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-search-information-retrieval","msr-locale-en_us"],"msr_publishername":"Association for Computational Linguistics.","msr_edition":"2015 Conference on Empirical Methods in Natural Language Processing, Lisbon, Portugal","msr_affiliation":"","msr_published_date":"2015-09-17","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"482\u2013491","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":"397859","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","title":"D15-1054","viewUrl":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-content\/uploads\/2017\/07\/D15-1054.pdf","id":397859,"label_id":0}],"msr_related_uploader":"","msr_citation_count":0,"msr_citation_count_updated":"","msr_s2_paper_id":"","msr_influential_citations":0,"msr_reference_count":0,"msr_arxiv_id":"","msr_s2_author_ids":[],"msr_s2_open_access":false,"msr_s2_pdf_url":null,"msr_attachments":[],"msr-author-ordering":[{"type":"user_nicename","value":"sule","user_id":36377,"rest_url":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=sule"},{"type":"text","value":"Yifan Hu","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[],"msr_project":[],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"inproceedings","related_content":[],"_links":{"self":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/397838","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":1,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/397838\/revisions"}],"predecessor-version":[{"id":397868,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/397838\/revisions\/397868"}],"wp:attachment":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/media?parent=397838"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=397838"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=397838"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=397838"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=397838"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=397838"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=397838"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=397838"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=397838"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=397838"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=397838"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=397838"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=397838"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}