{"id":"https://openalex.org/W4400374549","doi":"https://doi.org/10.1145/3637528.3671883","title":"Reliable Confidence Intervals for Information Retrieval Evaluation Using Generative A.I.","display_name":"Reliable Confidence Intervals for Information Retrieval Evaluation Using Generative A.I.","publication_year":2024,"publication_date":"2024-08-24","ids":{"openalex":"https://openalex.org/W4400374549","doi":"https://doi.org/10.1145/3637528.3671883"},"language":"en","primary_location":{"id":"doi:10.1145/3637528.3671883","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3637528.3671883","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3637528.3671883","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 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining","raw_type":"proceedings-article"},"type":"article","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3637528.3671883","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5002072527","display_name":"Harrie Oosterhuis","orcid":"https://orcid.org/0000-0002-0458-9233"},"institutions":[{"id":"https://openalex.org/I145872427","display_name":"Radboud University Nijmegen","ror":"https://ror.org/016xsfp80","country_code":"NL","type":"education","lineage":["https://openalex.org/I145872427"]}],"countries":["NL"],"is_corresponding":false,"raw_author_name":"Harrie Oosterhuis","raw_affiliation_strings":["Google Research & Radboud University, Amsterdam, Netherlands"],"raw_orcid":"https://orcid.org/0000-0002-0458-9233","affiliations":[{"raw_affiliation_string":"Google Research & Radboud University, Amsterdam, Netherlands","institution_ids":["https://openalex.org/I145872427"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5042795557","display_name":"Rolf Jagerman","orcid":"https://orcid.org/0000-0002-5169-495X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rolf Jagerman","raw_affiliation_strings":["Google Research, Amsterdam, Netherlands"],"raw_orcid":"https://orcid.org/0000-0002-5169-495X","affiliations":[{"raw_affiliation_string":"Google Research, Amsterdam, Netherlands","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100763094","display_name":"Zhen Qin","orcid":"https://orcid.org/0000-0001-6739-134X"},"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, New York City, USA"],"raw_orcid":"https://orcid.org/0000-0001-6739-134X","affiliations":[{"raw_affiliation_string":"Google Research, New York City, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"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, USA"],"raw_orcid":"https://orcid.org/0009-0000-1388-1423","affiliations":[{"raw_affiliation_string":"Google Research, Mountain View, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","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, USA"],"raw_orcid":"https://orcid.org/0000-0002-2941-6240","affiliations":[{"raw_affiliation_string":"Google Research, Mountain View, USA","institution_ids":["https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.6109,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.72578097,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":95,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"2307","last_page":"2317"},"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":0.9997000098228455,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9959999918937683,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/computer-science","display_name":"Computer science","score":0.8099195957183838},{"id":"https://openalex.org/keywords/relevance","display_name":"Relevance (law)","score":0.7122011184692383},{"id":"https://openalex.org/keywords/ranking","display_name":"Ranking (information retrieval)","score":0.659631073474884},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6082775592803955},{"id":"https://openalex.org/keywords/bootstrapping","display_name":"Bootstrapping (finance)","score":0.5991920232772827},{"id":"https://openalex.org/keywords/information-retrieval","display_name":"Information retrieval","score":0.5507491827011108},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.5379624366760254},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.49360764026641846},{"id":"https://openalex.org/keywords/data-mining","display_name":"Data mining","score":0.4542641043663025},{"id":"https://openalex.org/keywords/annotation","display_name":"Annotation","score":0.451631635427475},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.41740691661834717}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8099195957183838},{"id":"https://openalex.org/C158154518","wikidata":"https://www.wikidata.org/wiki/Q7310970","display_name":"Relevance (law)","level":2,"score":0.7122011184692383},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.659631073474884},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6082775592803955},{"id":"https://openalex.org/C207609745","wikidata":"https://www.wikidata.org/wiki/Q4944086","display_name":"Bootstrapping (finance)","level":2,"score":0.5991920232772827},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.5507491827011108},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5379624366760254},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.49360764026641846},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.4542641043663025},{"id":"https://openalex.org/C2776321320","wikidata":"https://www.wikidata.org/wiki/Q857525","display_name":"Annotation","level":2,"score":0.451631635427475},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41740691661834717},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"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/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C199539241","wikidata":"https://www.wikidata.org/wiki/Q7748","display_name":"Law","level":1,"score":0.0},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.0},{"id":"https://openalex.org/C106159729","wikidata":"https://www.wikidata.org/wiki/Q2294553","display_name":"Financial economics","level":1,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1145/3637528.3671883","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3637528.3671883","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3637528.3671883","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 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2407.02464","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2407.02464","pdf_url":"https://arxiv.org/pdf/2407.02464","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:oai:repository.ubn.ru.nl:2066/311017","is_oa":true,"landing_page_url":"https://hdl.handle.net/2066/311017","pdf_url":"https://repository.ubn.ru.nl//bitstream/handle/2066/311017/311017.pdf","source":{"id":"https://openalex.org/S4306401067","display_name":"Radboud Repository (Radboud University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I145872427","host_organization_name":"Radboud University Nijmegen","host_organization_lineage":["https://openalex.org/I145872427"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"Article in monograph or in proceedings"}],"best_oa_location":{"id":"doi:10.1145/3637528.3671883","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3637528.3671883","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3637528.3671883","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 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Decent work and economic growth","id":"https://metadata.un.org/sdg/8","score":0.5299999713897705}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4400374549.pdf"},"referenced_works_count":44,"referenced_works":["https://openalex.org/W79139011","https://openalex.org/W1490022694","https://openalex.org/W1521746852","https://openalex.org/W1523019309","https://openalex.org/W1981208470","https://openalex.org/W1985554184","https://openalex.org/W1990190154","https://openalex.org/W2007815473","https://openalex.org/W2008285050","https://openalex.org/W2009954908","https://openalex.org/W2015338694","https://openalex.org/W2045584315","https://openalex.org/W2052569738","https://openalex.org/W2066289596","https://openalex.org/W2069870183","https://openalex.org/W2111310810","https://openalex.org/W2115536324","https://openalex.org/W2124504084","https://openalex.org/W2126728600","https://openalex.org/W2156296249","https://openalex.org/W2181251292","https://openalex.org/W2291975472","https://openalex.org/W2514569044","https://openalex.org/W2613889334","https://openalex.org/W2765811365","https://openalex.org/W2774514250","https://openalex.org/W2798663050","https://openalex.org/W2922386288","https://openalex.org/W3005026984","https://openalex.org/W3096831136","https://openalex.org/W3133702157","https://openalex.org/W3155375847","https://openalex.org/W4206396334","https://openalex.org/W4206601343","https://openalex.org/W4226278401","https://openalex.org/W4244536387","https://openalex.org/W4284669679","https://openalex.org/W4284689925","https://openalex.org/W4288280764","https://openalex.org/W4297325261","https://openalex.org/W4313678819","https://openalex.org/W4318718892","https://openalex.org/W4385688511","https://openalex.org/W4401043313"],"related_works":["https://openalex.org/W1534274833","https://openalex.org/W3117246195","https://openalex.org/W2361861616","https://openalex.org/W156620619","https://openalex.org/W2616249226","https://openalex.org/W2098233217","https://openalex.org/W2914363205","https://openalex.org/W2997844990","https://openalex.org/W1598221548","https://openalex.org/W2081850291"],"abstract_inverted_index":{"The":[0],"traditional":[1],"evaluation":[2,55,82],"of":[3],"information":[4],"retrieval":[5],"(IR)":[6],"systems":[7],"is":[8],"generally":[9],"very":[10],"costly":[11],"as":[12,75],"it":[13,58],"requires":[14],"manual":[15],"relevance":[16,34,66],"annotation":[17],"from":[18],"human":[19],"experts.":[20],"Recent":[21],"advancements":[22],"in":[23],"generative":[24],"artificial":[25],"intelligence":[26],"-specifically":[27],"large":[28],"language":[29],"models":[30],"(LLMs)-":[31],"can":[32],"generate":[33],"annotations":[35,67],"at":[36],"an":[37],"enormous":[38],"scale":[39],"with":[40,53],"relatively":[41],"small":[42],"computational":[43],"costs.":[44],"Potentially,":[45],"this":[46],"could":[47],"alleviate":[48],"the":[49],"costs":[50],"traditionally":[51],"associated":[52],"IR":[54],"and":[56,74],"make":[57],"applicable":[59],"to":[60,71],"numerous":[61],"low-resource":[62],"applications.":[63],"However,":[64],"generated":[65],"are":[68],"not":[69],"immune":[70],"(systematic)":[72],"errors,":[73],"a":[76],"result,":[77],"directly":[78],"using":[79],"them":[80],"for":[81],"produces":[83],"unreliable":[84],"results.":[85]},"counts_by_year":[{"year":2025,"cited_by_count":2}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}