{"id":"https://openalex.org/W3022329361","doi":"https://doi.org/10.18653/v1/2020.acl-main.242","title":"Posterior Calibrated Training on Sentence Classification Tasks","display_name":"Posterior Calibrated Training on Sentence Classification Tasks","publication_year":2020,"publication_date":"2020-01-01","ids":{"openalex":"https://openalex.org/W3022329361","doi":"https://doi.org/10.18653/v1/2020.acl-main.242","mag":"3022329361"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2020.acl-main.242","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2020.acl-main.242","pdf_url":"https://www.aclweb.org/anthology/2020.acl-main.242.pdf","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 58th Annual Meeting of the Association for Computational Linguistics","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.aclweb.org/anthology/2020.acl-main.242.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5011213171","display_name":"Taehee Jung","orcid":null},"institutions":[{"id":"https://openalex.org/I170201317","display_name":"University of Pittsburgh","ror":"https://ror.org/01an3r305","country_code":"US","type":"education","lineage":["https://openalex.org/I170201317"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Taehee Jung","raw_affiliation_strings":["Department of Statistics, University of Pittsburgh, Pittsburgh, PA, USA","Univ. of Pittsburgh"],"affiliations":[{"raw_affiliation_string":"Department of Statistics, University of Pittsburgh, Pittsburgh, PA, USA","institution_ids":["https://openalex.org/I170201317"]},{"raw_affiliation_string":"Univ. of Pittsburgh","institution_ids":["https://openalex.org/I170201317"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040821714","display_name":"Dongyeop Kang","orcid":"https://orcid.org/0000-0002-9021-1789"},"institutions":[{"id":"https://openalex.org/I170201317","display_name":"University of Pittsburgh","ror":"https://ror.org/01an3r305","country_code":"US","type":"education","lineage":["https://openalex.org/I170201317"]},{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dongyeop Kang","raw_affiliation_strings":["School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA","Univ. of Pittsburgh"],"affiliations":[{"raw_affiliation_string":"School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA","institution_ids":["https://openalex.org/I74973139"]},{"raw_affiliation_string":"Univ. of Pittsburgh","institution_ids":["https://openalex.org/I170201317"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102840416","display_name":"Hua Cheng","orcid":"https://orcid.org/0000-0003-1109-7832"},"institutions":[{"id":"https://openalex.org/I4210142933","display_name":"Clinicas de Salud del Pueblo","ror":"https://ror.org/052rn8031","country_code":"US","type":"healthcare","lineage":["https://openalex.org/I4210142933"]},{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hua Cheng","raw_affiliation_strings":["3M|M*Modal, Pittsburgh, PA, USA","Carnegie Mellon University"],"affiliations":[{"raw_affiliation_string":"3M|M*Modal, Pittsburgh, PA, USA","institution_ids":["https://openalex.org/I4210142933"]},{"raw_affiliation_string":"Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5059204106","display_name":"Lucas Mentch","orcid":"https://orcid.org/0000-0002-8983-0320"},"institutions":[{"id":"https://openalex.org/I170201317","display_name":"University of Pittsburgh","ror":"https://ror.org/01an3r305","country_code":"US","type":"education","lineage":["https://openalex.org/I170201317"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Lucas Mentch","raw_affiliation_strings":["Department of Statistics, University of Pittsburgh, Pittsburgh, PA, USA","University of Pittsburgh, Pittsburgh, United States"],"affiliations":[{"raw_affiliation_string":"Department of Statistics, University of Pittsburgh, Pittsburgh, PA, USA","institution_ids":["https://openalex.org/I170201317"]},{"raw_affiliation_string":"University of Pittsburgh, Pittsburgh, United States","institution_ids":["https://openalex.org/I170201317"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5010424125","display_name":"Thomas Schaaf","orcid":"https://orcid.org/0000-0002-9569-4759"},"institutions":[{"id":"https://openalex.org/I4210142933","display_name":"Clinicas de Salud del Pueblo","ror":"https://ror.org/052rn8031","country_code":"US","type":"healthcare","lineage":["https://openalex.org/I4210142933"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Thomas Schaaf","raw_affiliation_strings":["3M|M*Modal, Pittsburgh, PA, USA"],"affiliations":[{"raw_affiliation_string":"3M|M*Modal, Pittsburgh, PA, USA","institution_ids":["https://openalex.org/I4210142933"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5011213171"],"corresponding_institution_ids":["https://openalex.org/I170201317"],"apc_list":null,"apc_paid":null,"fwci":0.2743,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.62859021,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":94},"biblio":{"volume":null,"issue":null,"first_page":"2723","last_page":"2730"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.9983000159263611,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.9983000159263611,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.9976000189781189,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.9961000084877014,"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/posterior-probability","display_name":"Posterior probability","score":0.7678077220916748},{"id":"https://openalex.org/keywords/calibration","display_name":"Calibration","score":0.6870499849319458},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6721768379211426},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.5642386674880981},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5354255437850952},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4591194689273834},{"id":"https://openalex.org/keywords/sentence","display_name":"Sentence","score":0.4430219233036041},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.4057276248931885},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3394000828266144},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.31991660594940186},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.3153952956199646},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.23574334383010864}],"concepts":[{"id":"https://openalex.org/C57830394","wikidata":"https://www.wikidata.org/wiki/Q278079","display_name":"Posterior probability","level":3,"score":0.7678077220916748},{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.6870499849319458},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6721768379211426},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.5642386674880981},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5354255437850952},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4591194689273834},{"id":"https://openalex.org/C2777530160","wikidata":"https://www.wikidata.org/wiki/Q41796","display_name":"Sentence","level":2,"score":0.4430219233036041},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4057276248931885},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3394000828266144},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.31991660594940186},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.3153952956199646},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.23574334383010864},{"id":"https://openalex.org/C127313418","wikidata":"https://www.wikidata.org/wiki/Q1069","display_name":"Geology","level":0,"score":0.0},{"id":"https://openalex.org/C111368507","wikidata":"https://www.wikidata.org/wiki/Q43518","display_name":"Oceanography","level":1,"score":0.0},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.18653/v1/2020.acl-main.242","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2020.acl-main.242","pdf_url":"https://www.aclweb.org/anthology/2020.acl-main.242.pdf","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 58th Annual Meeting of the Association for Computational Linguistics","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2004.14500","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2004.14500","pdf_url":"https://arxiv.org/pdf/2004.14500","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"text"},{"id":"mag:3022329361","is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2004.14500.pdf","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},{"id":"doi:10.48550/arxiv.2004.14500","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2004.14500","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.18653/v1/2020.acl-main.242","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2020.acl-main.242","pdf_url":"https://www.aclweb.org/anthology/2020.acl-main.242.pdf","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 58th Annual Meeting of the Association for Computational Linguistics","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.6600000262260437,"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3022329361.pdf","grobid_xml":"https://content.openalex.org/works/W3022329361.grobid-xml"},"referenced_works_count":13,"referenced_works":["https://openalex.org/W1598033630","https://openalex.org/W1618905105","https://openalex.org/W1775813496","https://openalex.org/W2012942264","https://openalex.org/W2135046866","https://openalex.org/W2803277810","https://openalex.org/W2963266575","https://openalex.org/W2963310665","https://openalex.org/W2963341956","https://openalex.org/W2964031043","https://openalex.org/W2964212410","https://openalex.org/W2985976073","https://openalex.org/W2990037398"],"related_works":["https://openalex.org/W3034704352","https://openalex.org/W2541906951","https://openalex.org/W3028500377","https://openalex.org/W2127824378","https://openalex.org/W2155045454","https://openalex.org/W3213809193","https://openalex.org/W2109045727","https://openalex.org/W2096258430","https://openalex.org/W2963747784","https://openalex.org/W2368526918","https://openalex.org/W3108002931","https://openalex.org/W2093478938","https://openalex.org/W2963727906","https://openalex.org/W3196611770","https://openalex.org/W2162432783","https://openalex.org/W3003859915","https://openalex.org/W2977330048","https://openalex.org/W3206550936","https://openalex.org/W3191747585","https://openalex.org/W3179123687"],"abstract_inverted_index":{"Most":[0],"classification":[1,183],"models":[2],"work":[3],"by":[4,124],"first":[5],"predicting":[6],"a":[7,186],"posterior":[8,31,65,88,106],"probability":[9,32],"distribution":[10],"over":[11],"all":[12],"classes":[13],"and":[14,104,141,163],"then":[15],"selecting":[16],"that":[17,92,110,196],"class":[18,47],"with":[19,155],"the":[20,28,44,69,95,99,102,116,151,169,194,203,207],"largest":[21],"estimated":[22],"probability.":[23],"In":[24],"many":[25],"settings":[26],"however,":[27],"quality":[29],"of":[30,129,137,143,182,188,213],"itself":[33],"(e.g.,":[34],"65%":[35],"chance":[36],"having":[37],"diabetes),":[38],"gives":[39],"more":[40],"reliable":[41],"information":[42],"than":[43],"final":[45,78],"predicted":[46,70,103],"alone.":[48],"When":[49],"these":[50],"methods":[51],"are":[52],"shown":[53],"to":[54,60,179],"be":[55,176],"poorly":[56],"calibrated,":[57],"most":[58],"fixes":[59],"date":[61],"have":[62],"relied":[63],"on":[64,77,147,160],"calibration,":[66],"which":[67],"rescales":[68],"probabilities":[71],"but":[72,119,166],"often":[73],"has":[74,193],"little":[75],"impact":[76],"classifications.":[79],"Here":[80],"we":[81],"propose":[82],"an":[83],"end-to-end":[84],"training":[85,91,174,208,215],"procedure":[86],"called":[87],"calibrated":[89],"(PosCal)":[90],"directly":[93],"optimizes":[94],"objective":[96,205],"while":[97],"minimizing":[98],"difference":[100],"between":[101],"empirical":[105],"probabilities.":[107],"We":[108,149],"show":[109],"PosCal":[111,133,173,192],"not":[112,167],"only":[113],"helps":[114],"reduce":[115],"calibration":[117,144,157,171,204],"error":[118,145,158],"also":[120],"improve":[121],"task":[122,138,153],"performance":[123,128,139,154],"penalizing":[125],"drops":[126],"in":[127],"both":[130],"objectives.":[131],"Our":[132],"achieves":[134],"about":[135],"2.5%":[136],"gain":[140],"16.1%":[142],"reduction":[146,159],"GLUE":[148],"achieved":[150],"comparable":[152],"13.2%":[156],"xSLUE":[161],"(Kang":[162],"Hovy,":[164],"2019),":[165],"outperforming":[168],"two-stage":[170],"baseline.":[172],"can":[175],"easily":[177],"extendable":[178],"any":[180],"types":[181],"tasks":[184],"as":[185],"form":[187],"regularization":[189],"term.":[190],"Also,":[191],"advantage":[195],"it":[197],"incrementally":[198],"tracks":[199],"needed":[200],"statistics":[201],"for":[202],"during":[206],"process,":[209],"making":[210],"efficient":[211],"use":[212],"large":[214],"sets":[216],"1":[217],".":[218]},"counts_by_year":[{"year":2022,"cited_by_count":1},{"year":2021,"cited_by_count":1}],"updated_date":"2026-03-10T16:38:18.471706","created_date":"2025-10-10T00:00:00"}