{"id":335009,"date":"2016-12-11T21:29:26","date_gmt":"2016-12-12T05:29:26","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-research-item&#038;p=335009"},"modified":"2018-10-16T20:07:01","modified_gmt":"2018-10-17T03:07:01","slug":"recstore-extensible-adaptive-framework-online-recommender-queries-inside-database-engine","status":"publish","type":"msr-research-item","link":"https:\/\/www.microsoft.com\/en-us\/research\/publication\/recstore-extensible-adaptive-framework-online-recommender-queries-inside-database-engine\/","title":{"rendered":"RecStore: an extensible and adaptive framework for online recommender queries inside the database engine"},"content":{"rendered":"<div id=\"citationdetails\" class=\" x-tabs-top\">\n<div id=\"tab-body9\" class=\" x-tabs-body\">\n<div id=\"abstract\" class=\"ytab x-tabs-item-body\">\n<div class=\"tabbody\">\n<div>\n<p>Most recommendation methods (e.g., collaborative filtering) consist of (1) a computationally intense <i>offline<\/i> phase that computes a recommender model based on users&#8217; opinions of items, and (2) an <i>online<\/i> phase consisting of SQL-based queries that use the model (generated offline) to derive user preferences and provide recommendations for interesting items. Current application usage trends require a completely <i>online<\/i> recommender process, meaning the recommender model must update in <i>real time<\/i> as new opinions enter the system. To tackle this problem, we propose <i>RecStore<\/i>, a DBMS storage engine module capable of efficient <i>online model maintenance<\/i>. Externally, models managed by <i>RecStore<\/i> behave as relational tables, thus <i>existing<\/i> SQL-based recommendation queries remain <i>unchanged<\/i> while gaining online model support. <i>RecStore<\/i> maintains internal statistics and data structures aimed at providing efficient incremental updates to the recommender model, while employing an <i>adaptive<\/i> strategy for internal maintenance and load shedding to realize a balance between efficiency in updates or query processing based on system workloads. <i>RecStore<\/i> is also <i>extensible<\/i>, supporting a declarative syntax for defining recommender models. The efficacy of <i>RecStore<\/i> is demonstrated by providing the implementation details of three state-of-the-art collaborative filtering models. We provide an extensive experimental evaluation of a prototype of <i>RecStore<\/i>, built inside the storage engine of PostgreSQL, using a real-life recommender system workload.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Most recommendation methods (e.g., collaborative filtering) consist of (1) a computationally intense offline phase that computes a recommender model based on users&#8217; opinions of items, and (2) an online phase consisting of SQL-based queries that use the model (generated offline) to derive user preferences and provide recommendations for interesting items. Current application usage trends require [&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":"ACM","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"EDBT '12 Proceedings of the 15th International Conference on Extending Database Technology","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"","msr_journal":"","msr_number":"","msr_organization":"","msr_pages_string":"","msr_page_range_start":"","msr_page_range_end":"","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"EDBT '12 Proceedings of the 15th International Conference on Extending Database Technology","msr_doi":"10.1145\/2247596.2247608","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":"2012-03-27","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"http:\/\/dl.acm.org\/citation.cfm?id=2247608","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":[13563],"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-335009","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-data-platform-analytics","msr-locale-en_us"],"msr_publishername":"ACM","msr_edition":"EDBT '12 Proceedings of the 15th International Conference on Extending Database Technology","msr_affiliation":"","msr_published_date":"2012-03-27","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":"http:\/\/dl.acm.org\/citation.cfm?id=2247608","msr_doi":"10.1145\/2247596.2247608","msr_publication_uploader":[{"type":"url","title":"http:\/\/dl.acm.org\/citation.cfm?id=2247608","viewUrl":false,"id":false,"label_id":0},{"type":"doi","title":"10.1145\/2247596.2247608","viewUrl":false,"id":false,"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":[{"id":0,"url":"http:\/\/dl.acm.org\/citation.cfm?id=2247608"}],"msr-author-ordering":[{"type":"user_nicename","value":"justinle","user_id":32451,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=justinle"},{"type":"text","value":"Mohamed Sarwat","user_id":0,"rest_url":false},{"type":"text","value":"Mohamed F. Mokbel","user_id":0,"rest_url":false},{"type":"text","value":"Michael D. Ekstrand","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:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/335009","targetHints":{"allow":["GET"]}}],"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\/335009\/revisions"}],"predecessor-version":[{"id":522853,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/335009\/revisions\/522853"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=335009"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=335009"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=335009"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=335009"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=335009"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=335009"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=335009"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=335009"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=335009"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=335009"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=335009"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=335009"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=335009"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}