{"id":"https://openalex.org/W3189117283","doi":"https://doi.org/10.1145/3471158.3472238","title":"Ensemble Distillation for BERT-Based Ranking Models","display_name":"Ensemble Distillation for BERT-Based Ranking Models","publication_year":2021,"publication_date":"2021-07-11","ids":{"openalex":"https://openalex.org/W3189117283","doi":"https://doi.org/10.1145/3471158.3472238","mag":"3189117283"},"language":"en","primary_location":{"id":"doi:10.1145/3471158.3472238","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3471158.3472238","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3471158.3472238","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2021 ACM SIGIR International Conference on Theory of Information Retrieval","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3471158.3472238","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5011279860","display_name":"Honglei Zhuang","orcid":"https://orcid.org/0000-0001-8134-1509"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Honglei Zhuang","raw_affiliation_strings":["Google Research, Mountain View, CA, USA"],"affiliations":[{"raw_affiliation_string":"Google Research, Mountain View, CA, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100763095","display_name":"Zhen Qin","orcid":"https://orcid.org/0000-0001-7857-9719"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhen Qin","raw_affiliation_strings":["Google Research, Mountain View, CA, USA"],"affiliations":[{"raw_affiliation_string":"Google Research, Mountain View, CA, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071179467","display_name":"Shuguang Han","orcid":"https://orcid.org/0000-0003-1416-6960"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shuguang Han","raw_affiliation_strings":["Alibaba, Hangzhou, China"],"affiliations":[{"raw_affiliation_string":"Alibaba, Hangzhou, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064608039","display_name":"Xuanhui Wang","orcid":"https://orcid.org/0009-0000-1388-1423"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xuanhui Wang","raw_affiliation_strings":["Google Research, Mountain View, CA, USA"],"affiliations":[{"raw_affiliation_string":"Google Research, Mountain View, CA, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032248436","display_name":"Michael Bendersky","orcid":"https://orcid.org/0000-0002-2941-6240"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Michael Bendersky","raw_affiliation_strings":["Google Research, Mountain View, CA, USA"],"affiliations":[{"raw_affiliation_string":"Google Research, Mountain View, CA, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5037200145","display_name":"Marc Najork","orcid":"https://orcid.org/0000-0003-1423-0854"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Marc Najork","raw_affiliation_strings":["Google Research, Mountain View, CA, USA"],"affiliations":[{"raw_affiliation_string":"Google Research, Mountain View, CA, USA","institution_ids":["https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5011279860"],"corresponding_institution_ids":["https://openalex.org/I1291425158"],"apc_list":null,"apc_paid":null,"fwci":1.6316,"has_fulltext":true,"cited_by_count":15,"citation_normalized_percentile":{"value":0.86616853,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"131","last_page":"136"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9998000264167786,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9993000030517578,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11550","display_name":"Text and Document Classification Technologies","score":0.9941999912261963,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/ranking","display_name":"Ranking (information retrieval)","score":0.8371962308883667},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7467924952507019},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.7081711292266846},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.6634634733200073},{"id":"https://openalex.org/keywords/ensemble-forecasting","display_name":"Ensemble forecasting","score":0.6304600834846497},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6254865527153015},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.5577347278594971},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5108489394187927},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.46688833832740784},{"id":"https://openalex.org/keywords/ensemble-learning","display_name":"Ensemble learning","score":0.46354764699935913},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.4162990152835846},{"id":"https://openalex.org/keywords/bootstrap-aggregating","display_name":"Bootstrap aggregating","score":0.41234761476516724},{"id":"https://openalex.org/keywords/data-mining","display_name":"Data mining","score":0.35142993927001953}],"concepts":[{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.8371962308883667},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7467924952507019},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.7081711292266846},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.6634634733200073},{"id":"https://openalex.org/C119898033","wikidata":"https://www.wikidata.org/wiki/Q3433888","display_name":"Ensemble forecasting","level":2,"score":0.6304600834846497},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6254865527153015},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5577347278594971},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5108489394187927},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.46688833832740784},{"id":"https://openalex.org/C45942800","wikidata":"https://www.wikidata.org/wiki/Q245652","display_name":"Ensemble learning","level":2,"score":0.46354764699935913},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.4162990152835846},{"id":"https://openalex.org/C162040801","wikidata":"https://www.wikidata.org/wiki/Q799897","display_name":"Bootstrap aggregating","level":2,"score":0.41234761476516724},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.35142993927001953},{"id":"https://openalex.org/C178790620","wikidata":"https://www.wikidata.org/wiki/Q11351","display_name":"Organic chemistry","level":1,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C121955636","wikidata":"https://www.wikidata.org/wiki/Q4116214","display_name":"Accounting","level":1,"score":0.0},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3471158.3472238","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3471158.3472238","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3471158.3472238","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2021 ACM SIGIR International Conference on Theory of Information Retrieval","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3471158.3472238","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3471158.3472238","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3471158.3472238","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2021 ACM SIGIR International Conference on Theory of Information Retrieval","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.4699999988079071,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3189117283.pdf","grobid_xml":"https://content.openalex.org/works/W3189117283.grobid-xml"},"referenced_works_count":26,"referenced_works":["https://openalex.org/W569478347","https://openalex.org/W2108862644","https://openalex.org/W2148972377","https://openalex.org/W2808847742","https://openalex.org/W2892181857","https://openalex.org/W2902365885","https://openalex.org/W2909544278","https://openalex.org/W2913668833","https://openalex.org/W2937408455","https://openalex.org/W2951534261","https://openalex.org/W2959353218","https://openalex.org/W2982596739","https://openalex.org/W2996834012","https://openalex.org/W3017018726","https://openalex.org/W3021052948","https://openalex.org/W3021397474","https://openalex.org/W3034870558","https://openalex.org/W3035313607","https://openalex.org/W3045033475","https://openalex.org/W3091207065","https://openalex.org/W3092952717","https://openalex.org/W3105107530","https://openalex.org/W3105136066","https://openalex.org/W3146365155","https://openalex.org/W3171713913","https://openalex.org/W4246571544"],"related_works":["https://openalex.org/W2794896638","https://openalex.org/W2891633941","https://openalex.org/W3202800081","https://openalex.org/W3100759197","https://openalex.org/W2411183043","https://openalex.org/W3101614107","https://openalex.org/W1909207154","https://openalex.org/W3036530763","https://openalex.org/W3130261933","https://openalex.org/W2791865700"],"abstract_inverted_index":{"Over":[0],"the":[1,44,49,70,82,90,99,133,139,153,159,163,170,195],"past":[2],"two":[3],"years,":[4],"large":[5],"pretrained":[6],"language":[7],"models":[8,39,56,88,174,180],"such":[9],"as":[10,119,121],"BERT":[11],"have":[12],"been":[13],"applied":[14],"to":[15,69,80,192],"text":[16],"ranking":[17,38,106,127,137,203],"problems":[18],"and":[19,138],"showed":[20],"superior":[21],"performance":[22,50,83,160],"on":[23,132,182,194],"multiple":[24,36,122,166],"public":[25],"benchmark":[26],"data":[27,141],"sets.":[28],"Prior":[29],"work":[30],"demonstrated":[31],"that":[32,146],"an":[33,53,85,199],"ensemble":[34,54,86,100,164,200],"of":[35,55,72,84,87,93,113,162,165,172,201],"BERT-based":[37,104,202],"can":[40,156],"not":[41],"only":[42],"boost":[43],"performance,":[45],"but":[46],"also":[47,176],"reduce":[48],"variance.":[51],"However,":[52],"is":[57],"more":[58,177],"costly":[59],"because":[60],"it":[61],"needs":[62],"computing":[63],"resource":[64],"and/or":[65],"inference":[66,91],"time":[67],"proportional":[68],"number":[71],"models.":[73,167,204],"In":[74],"this":[75],"paper,":[76],"we":[77,109],"study":[78,110],"how":[79],"retain":[81,158],"at":[89],"cost":[92],"a":[94,102,189],"single":[95,103],"model":[96,155],"by":[97,198],"distilling":[98],"into":[101],"student":[105],"model.":[107],"Specifically,":[108],"different":[111],"designs":[112],"teacher":[114],"labels,":[115],"various":[116],"distillation":[117,123,151],"strategies,":[118],"well":[120],"losses":[124],"tailored":[125],"for":[126],"problems.":[128],"We":[129],"conduct":[130],"experiments":[131],"MS":[134],"MARCO":[135],"passage":[136],"TREC-COVID":[140],"set.":[142],"Our":[143],"results":[144,187],"show":[145],"even":[147],"with":[148],"these":[149],"simple":[150],"techniques,":[152],"distilled":[154,173],"effectively":[157],"gain":[161],"More":[168],"interestingly,":[169],"performances":[171],"are":[175],"stable":[178],"than":[179],"fine-tuned":[181],"original":[183],"labeled":[184],"data.":[185],"The":[186],"reveal":[188],"promising":[190],"direction":[191],"capitalize":[193],"gains":[196],"achieved":[197]},"counts_by_year":[{"year":2025,"cited_by_count":3},{"year":2024,"cited_by_count":4},{"year":2023,"cited_by_count":5},{"year":2022,"cited_by_count":2},{"year":2021,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}