{"id":"https://openalex.org/W2896780650","doi":"https://doi.org/10.18653/v1/p18-1060","title":"The price of debiasing automatic metrics in natural language evalaution","display_name":"The price of debiasing automatic metrics in natural language evalaution","publication_year":2018,"publication_date":"2018-01-01","ids":{"openalex":"https://openalex.org/W2896780650","doi":"https://doi.org/10.18653/v1/p18-1060","mag":"2896780650"},"language":"en","primary_location":{"id":"doi:10.18653/v1/p18-1060","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/p18-1060","pdf_url":"https://www.aclweb.org/anthology/P18-1060.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 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.aclweb.org/anthology/P18-1060.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5024150057","display_name":"Arun Tejasvi Chaganty","orcid":"https://orcid.org/0000-0001-7122-1298"},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Arun Chaganty","raw_affiliation_strings":["Computer Science Department, Stanford University"],"affiliations":[{"raw_affiliation_string":"Computer Science Department, Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034715099","display_name":"Stephen Mussmann","orcid":null},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Stephen Mussmann","raw_affiliation_strings":["Computer Science Department, Stanford University"],"affiliations":[{"raw_affiliation_string":"Computer Science Department, Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5025255782","display_name":"Percy Liang","orcid":"https://orcid.org/0000-0002-0458-6139"},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Percy Liang","raw_affiliation_strings":["Computer Science Department, Stanford University"],"affiliations":[{"raw_affiliation_string":"Computer Science Department, Stanford University","institution_ids":["https://openalex.org/I97018004"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5024150057"],"corresponding_institution_ids":["https://openalex.org/I97018004"],"apc_list":null,"apc_paid":null,"fwci":12.1857,"has_fulltext":true,"cited_by_count":101,"citation_normalized_percentile":{"value":0.9875668,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"643","last_page":"653"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":1.0,"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":1.0,"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":0.9995999932289124,"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/T12031","display_name":"Speech and dialogue systems","score":0.9984999895095825,"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/automatic-summarization","display_name":"Automatic summarization","score":0.7900852560997009},{"id":"https://openalex.org/keywords/debiasing","display_name":"Debiasing","score":0.7885317802429199},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7313764691352844},{"id":"https://openalex.org/keywords/estimator","display_name":"Estimator","score":0.6712151169776917},{"id":"https://openalex.org/keywords/bias-of-an-estimator","display_name":"Bias of an estimator","score":0.6366640329360962},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.5883198976516724},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4013761579990387},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.37018316984176636},{"id":"https://openalex.org/keywords/minimum-variance-unbiased-estimator","display_name":"Minimum-variance unbiased estimator","score":0.2565377652645111},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.23565784096717834},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.15453600883483887}],"concepts":[{"id":"https://openalex.org/C170858558","wikidata":"https://www.wikidata.org/wiki/Q1394144","display_name":"Automatic summarization","level":2,"score":0.7900852560997009},{"id":"https://openalex.org/C2779458634","wikidata":"https://www.wikidata.org/wiki/Q24963715","display_name":"Debiasing","level":2,"score":0.7885317802429199},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7313764691352844},{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.6712151169776917},{"id":"https://openalex.org/C191393472","wikidata":"https://www.wikidata.org/wiki/Q15222032","display_name":"Bias of an estimator","level":4,"score":0.6366640329360962},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.5883198976516724},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4013761579990387},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.37018316984176636},{"id":"https://openalex.org/C165646398","wikidata":"https://www.wikidata.org/wiki/Q3755281","display_name":"Minimum-variance unbiased estimator","level":3,"score":0.2565377652645111},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.23565784096717834},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.15453600883483887},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.0},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.0},{"id":"https://openalex.org/C188147891","wikidata":"https://www.wikidata.org/wiki/Q147638","display_name":"Cognitive science","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":1,"locations":[{"id":"doi:10.18653/v1/p18-1060","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/p18-1060","pdf_url":"https://www.aclweb.org/anthology/P18-1060.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 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/p18-1060","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/p18-1060","pdf_url":"https://www.aclweb.org/anthology/P18-1060.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 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Decent work and economic growth","id":"https://metadata.un.org/sdg/8","score":0.49000000953674316}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2896780650.pdf","grobid_xml":"https://content.openalex.org/works/W2896780650.grobid-xml"},"referenced_works_count":41,"referenced_works":["https://openalex.org/W1544827683","https://openalex.org/W1861492603","https://openalex.org/W1956340063","https://openalex.org/W2030265142","https://openalex.org/W2070150502","https://openalex.org/W2101105183","https://openalex.org/W2108682071","https://openalex.org/W2117010802","https://openalex.org/W2119196781","https://openalex.org/W2123442489","https://openalex.org/W2133459682","https://openalex.org/W2149327368","https://openalex.org/W2295951612","https://openalex.org/W2341401723","https://openalex.org/W2469104253","https://openalex.org/W2525778437","https://openalex.org/W2558203065","https://openalex.org/W2593751037","https://openalex.org/W2605035112","https://openalex.org/W2606974598","https://openalex.org/W2612675303","https://openalex.org/W2615953416","https://openalex.org/W2626154462","https://openalex.org/W2745039414","https://openalex.org/W2759567155","https://openalex.org/W2760672101","https://openalex.org/W2916445322","https://openalex.org/W2949615363","https://openalex.org/W2951493172","https://openalex.org/W2951534261","https://openalex.org/W2963173382","https://openalex.org/W2963527228","https://openalex.org/W2963672599","https://openalex.org/W2963768805","https://openalex.org/W2963863909","https://openalex.org/W2963903950","https://openalex.org/W2963929190","https://openalex.org/W2963963993","https://openalex.org/W4234807459","https://openalex.org/W6601289607","https://openalex.org/W6835070974"],"related_works":["https://openalex.org/W4362554880","https://openalex.org/W4386875279","https://openalex.org/W4281684980","https://openalex.org/W2171721708","https://openalex.org/W3214527415","https://openalex.org/W4287887864","https://openalex.org/W1495104519","https://openalex.org/W4390963114","https://openalex.org/W4225584739","https://openalex.org/W4224293420"],"abstract_inverted_index":{"For":[0],"evaluating":[1,83],"generation":[2],"systems,":[3],"automatic":[4,55,114],"metrics":[5,56],"such":[6],"as":[7],"BLEU":[8],"cost":[9,67,80,132],"nothing":[10],"to":[11,17,24,53,60,120,126,129],"run":[12],"but":[13],"have":[14],"been":[15],"shown":[16,119],"correlate":[18],"poorly":[19],"with":[20,57,65,103],"human":[21,36,58,69,121],"judgment,":[22],"leading":[23],"systematic":[25],"bias":[26],"against":[27],"certain":[28],"model":[29],"improvements.":[30],"On":[31],"the":[32,38,110,117],"other":[33],"hand,":[34],"averaging":[35],"judgments,":[37],"unbiased":[39,63,101],"gold":[40],"standard,":[41],"is":[42,96,99],"often":[43],"too":[44],"expensive.":[45],"In":[46,72],"this":[47],"paper,":[48],"we":[49,75],"use":[50],"control":[51],"variates":[52],"combine":[54],"evaluation":[59,70],"obtain":[61,76,130],"an":[62],"estimator":[64,95,102],"lower":[66,104],"than":[68],"alone.":[71],"practice,":[73],"however,":[74],"only":[77],"a":[78],"7-13%":[79],"reduction":[81],"on":[82],"summarization":[84],"and":[85,116],"open-response":[86],"question":[87],"answering":[88],"systems.":[89],"We":[90],"then":[91],"prove":[92],"that":[93],"our":[94],"optimal:":[97],"there":[98],"no":[100],"cost.":[105],"Our":[106],"theory":[107],"further":[108],"highlights":[109],"two":[111],"fundamental":[112],"bottlenecks-the":[113],"metric":[115],"prompt":[118],"evaluatorsboth":[122],"of":[123],"which":[124],"need":[125],"be":[127],"improved":[128],"greater":[131],"savings.":[133]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":9},{"year":2022,"cited_by_count":12},{"year":2021,"cited_by_count":30},{"year":2020,"cited_by_count":18},{"year":2019,"cited_by_count":24}],"updated_date":"2026-04-14T06:02:45.956762","created_date":"2025-10-10T00:00:00"}