{"id":"https://openalex.org/W4411630288","doi":"https://doi.org/10.18653/v1/2024.eacl-long.8","title":"LLM Comparative Assessment: Zero-shot NLG Evaluation through Pairwise Comparisons using Large Language Models","display_name":"LLM Comparative Assessment: Zero-shot NLG Evaluation through Pairwise Comparisons using Large Language Models","publication_year":2024,"publication_date":"2024-01-01","ids":{"openalex":"https://openalex.org/W4411630288","doi":"https://doi.org/10.18653/v1/2024.eacl-long.8"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2024.eacl-long.8","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2024.eacl-long.8","pdf_url":"https://aclanthology.org/2024.eacl-long.8.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 18th Conference of the European Chapter 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://aclanthology.org/2024.eacl-long.8.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5011907463","display_name":"Adian Liusie","orcid":null},"institutions":[{"id":"https://openalex.org/I4210096386","display_name":"Bridge University","ror":"https://ror.org/00cbm0437","country_code":"SS","type":"education","lineage":["https://openalex.org/I4210096386"]}],"countries":["SS"],"is_corresponding":false,"raw_author_name":"Adian Liusie","raw_affiliation_strings":["ALTA Institute , Department of Engineering , University of Cambridge"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ALTA Institute , Department of Engineering , University of Cambridge","institution_ids":["https://openalex.org/I4210096386"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024447363","display_name":"Potsawee Manakul","orcid":null},"institutions":[{"id":"https://openalex.org/I4210096386","display_name":"Bridge University","ror":"https://ror.org/00cbm0437","country_code":"SS","type":"education","lineage":["https://openalex.org/I4210096386"]}],"countries":["SS"],"is_corresponding":false,"raw_author_name":"Potsawee Manakul","raw_affiliation_strings":["ALTA Institute , Department of Engineering , University of Cambridge"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ALTA Institute , Department of Engineering , University of Cambridge","institution_ids":["https://openalex.org/I4210096386"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5050766679","display_name":"Mark Gales","orcid":"https://orcid.org/0000-0002-5311-8219"},"institutions":[{"id":"https://openalex.org/I4210096386","display_name":"Bridge University","ror":"https://ror.org/00cbm0437","country_code":"SS","type":"education","lineage":["https://openalex.org/I4210096386"]}],"countries":["SS"],"is_corresponding":false,"raw_author_name":"Mark Gales","raw_affiliation_strings":["ALTA Institute , Department of Engineering , University of Cambridge"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ALTA Institute , Department of Engineering , University of Cambridge","institution_ids":["https://openalex.org/I4210096386"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I4210096386"],"apc_list":null,"apc_paid":null,"fwci":5.5789,"has_fulltext":true,"cited_by_count":20,"citation_normalized_percentile":{"value":0.9651806,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"139","last_page":"151"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9796000123023987,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9796000123023987,"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/T10028","display_name":"Topic Modeling","score":0.9067000150680542,"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/pairwise-comparison","display_name":"Pairwise comparison","score":0.8056608438491821},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6524455547332764},{"id":"https://openalex.org/keywords/zero","display_name":"Zero (linguistics)","score":0.6113475561141968},{"id":"https://openalex.org/keywords/shot","display_name":"Shot (pellet)","score":0.5905133485794067},{"id":"https://openalex.org/keywords/one-shot","display_name":"One shot","score":0.5211798548698425},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5025746822357178},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.423860639333725},{"id":"https://openalex.org/keywords/linguistics","display_name":"Linguistics","score":0.11655700206756592},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.06709927320480347},{"id":"https://openalex.org/keywords/chemistry","display_name":"Chemistry","score":0.06608739495277405}],"concepts":[{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.8056608438491821},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6524455547332764},{"id":"https://openalex.org/C2780813799","wikidata":"https://www.wikidata.org/wiki/Q3274237","display_name":"Zero (linguistics)","level":2,"score":0.6113475561141968},{"id":"https://openalex.org/C2778344882","wikidata":"https://www.wikidata.org/wiki/Q278938","display_name":"Shot (pellet)","level":2,"score":0.5905133485794067},{"id":"https://openalex.org/C2992734406","wikidata":"https://www.wikidata.org/wiki/Q413267","display_name":"One shot","level":2,"score":0.5211798548698425},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5025746822357178},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.423860639333725},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.11655700206756592},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.06709927320480347},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.06608739495277405},{"id":"https://openalex.org/C78519656","wikidata":"https://www.wikidata.org/wiki/Q101333","display_name":"Mechanical engineering","level":1,"score":0.0},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0},{"id":"https://openalex.org/C178790620","wikidata":"https://www.wikidata.org/wiki/Q11351","display_name":"Organic chemistry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2024.eacl-long.8","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2024.eacl-long.8","pdf_url":"https://aclanthology.org/2024.eacl-long.8.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 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2024.eacl-long.8","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2024.eacl-long.8","pdf_url":"https://aclanthology.org/2024.eacl-long.8.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 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.75,"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320320273","display_name":"University of Cambridge","ror":"https://ror.org/013meh722"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4411630288.pdf","grobid_xml":"https://content.openalex.org/works/W4411630288.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2497720472","https://openalex.org/W4292659306","https://openalex.org/W3044321615","https://openalex.org/W2806221744","https://openalex.org/W2326937258","https://openalex.org/W394267150","https://openalex.org/W2773965352","https://openalex.org/W4294892107","https://openalex.org/W2357748469","https://openalex.org/W2392917037"],"abstract_inverted_index":{"Current":[0],"developments":[1],"in":[2,23,78,114,120,158],"large":[3],"language":[4,15,29],"models":[5],"(LLMs)":[6],"have":[7],"enabled":[8],"impressive":[9],"zero-shot":[10,54],"capabilities":[11],"across":[12],"various":[13],"natural":[14,28],"tasks.An":[16],"interesting":[17],"application":[18],"of":[19,27,51,69,122,125],"these":[20],"systems":[21],"is":[22,132,152],"the":[24,48,115,123],"automated":[25],"assessment":[26,62,72,103,131,151],"generation":[30],"(NLG),":[31],"a":[32,133],"highly":[33],"challenging":[34],"area":[35],"with":[36,165],"great":[37],"practical":[38],"benefit.In":[39],"this":[40],"paper,":[41],"we":[42,81,168,182],"explore":[43],"two":[44,92],"options":[45,93],"for":[46,53,139],"exploiting":[47],"emergent":[49],"abilities":[50],"LLMs":[52,171],"NLG":[55,79,140],"assessment:":[56],"absolute":[57,110],"score":[58],"prediction,":[59],"and":[60,117,136,148,157,181],"comparative":[61,71,102,130,150],"which":[63],"uses":[64],"relative":[65],"comparisons":[66],"between":[67],"pairs":[68],"candidates.Though":[70],"has":[73],"not":[74],"been":[75],"extensively":[76],"studied":[77],"assessment,":[80],"note":[82],"that":[83,128,170,186],"humans":[84],"often":[85,172],"find":[86],"it":[87],"more":[88],"intuitive":[89],"to":[90,109,154],"compare":[91],"rather":[94],"than":[95],"scoring":[96],"each":[97],"one":[98],"independently.This":[99],"work":[100],"examines":[101],"from":[104],"multiple":[105],"perspectives:":[106],"performance":[107,163],"compared":[108],"grading;":[111],"positional":[112,175],"biases":[113,176],"prompt;":[116],"efficient":[118],"ranking":[119],"terms":[121],"number":[124],"comparisons.We":[126],"illustrate":[127],"LLM":[129],"simple,":[134],"general":[135],"effective":[137],"approach":[138],"assessment.For":[141],"moderatesized":[142],"open-source":[143],"LLMs,":[144],"such":[145],"as":[146],"FlanT5":[147],"Llama2-chat,":[149],"superior":[153],"prompt":[155],"scoring,":[156],"many":[159],"cases":[160],"can":[161,187],"achieve":[162],"competitive":[164],"state-ofthe-art":[166],"methods.Additionally,":[167],"demonstrate":[169],"exhibit":[173],"strong":[174],"when":[177],"making":[178],"pairwise":[179],"comparisons,":[180],"propose":[183],"debiasing":[184],"methods":[185],"further":[188],"improve":[189],"performance.":[190]},"counts_by_year":[{"year":2026,"cited_by_count":7},{"year":2025,"cited_by_count":13}],"updated_date":"2026-07-10T07:45:09.275182","created_date":"2025-10-10T00:00:00"}