{"id":"https://openalex.org/W2741893366","doi":"https://doi.org/10.18653/v1/e17-1039","title":"Efficient Benchmarking of NLP APIs using Multi-armed Bandits","display_name":"Efficient Benchmarking of NLP APIs using Multi-armed Bandits","publication_year":2017,"publication_date":"2017-01-01","ids":{"openalex":"https://openalex.org/W2741893366","doi":"https://doi.org/10.18653/v1/e17-1039","mag":"2741893366"},"language":"en","primary_location":{"id":"doi:10.18653/v1/e17-1039","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/e17-1039","pdf_url":"https://www.aclweb.org/anthology/E17-1039.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 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.aclweb.org/anthology/E17-1039.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5081525024","display_name":"Gholamreza Haffari","orcid":"https://orcid.org/0000-0001-7326-8380"},"institutions":[{"id":"https://openalex.org/I56590836","display_name":"Monash University","ror":"https://ror.org/02bfwt286","country_code":"AU","type":"education","lineage":["https://openalex.org/I56590836"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Gholamreza Haffari","raw_affiliation_strings":["Faculty of Information Technology, Monash University Clayton, VICTORIA 3800, Australia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Information Technology, Monash University Clayton, VICTORIA 3800, Australia","institution_ids":["https://openalex.org/I56590836"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066094876","display_name":"Tuan Dung Tran","orcid":null},"institutions":[{"id":"https://openalex.org/I56590836","display_name":"Monash University","ror":"https://ror.org/02bfwt286","country_code":"AU","type":"education","lineage":["https://openalex.org/I56590836"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Tuan Dung Tran","raw_affiliation_strings":["Faculty of Information Technology, Monash University Clayton, VICTORIA 3800, Australia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Information Technology, Monash University Clayton, VICTORIA 3800, Australia","institution_ids":["https://openalex.org/I56590836"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5069487278","display_name":"Mark Carman","orcid":"https://orcid.org/0000-0001-6575-9737"},"institutions":[{"id":"https://openalex.org/I56590836","display_name":"Monash University","ror":"https://ror.org/02bfwt286","country_code":"AU","type":"education","lineage":["https://openalex.org/I56590836"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Mark Carman","raw_affiliation_strings":["Faculty of Information Technology, Monash University Clayton, VICTORIA 3800, Australia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Information Technology, Monash University Clayton, VICTORIA 3800, Australia","institution_ids":["https://openalex.org/I56590836"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I56590836"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"408","last_page":"416"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12101","display_name":"Advanced Bandit Algorithms Research","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T12101","display_name":"Advanced Bandit Algorithms Research","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T12072","display_name":"Machine Learning and Algorithms","score":0.9980999827384949,"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/T12761","display_name":"Data Stream Mining Techniques","score":0.9954000115394592,"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/benchmarking","display_name":"Benchmarking","score":0.9305815100669861},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8002023696899414},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.6584110260009766},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5988695025444031},{"id":"https://openalex.org/keywords/measure","display_name":"Measure (data warehouse)","score":0.5400159955024719},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.5257509350776672},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.4340782165527344},{"id":"https://openalex.org/keywords/data-mining","display_name":"Data mining","score":0.36569005250930786},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.36494553089141846},{"id":"https://openalex.org/keywords/information-retrieval","display_name":"Information retrieval","score":0.34293287992477417}],"concepts":[{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.9305815100669861},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8002023696899414},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.6584110260009766},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5988695025444031},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.5400159955024719},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5257509350776672},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.4340782165527344},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.36569005250930786},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.36494553089141846},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.34293287992477417},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0},{"id":"https://openalex.org/C162853370","wikidata":"https://www.wikidata.org/wiki/Q39809","display_name":"Marketing","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/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/e17-1039","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/e17-1039","pdf_url":"https://www.aclweb.org/anthology/E17-1039.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 15th 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/e17-1039","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/e17-1039","pdf_url":"https://www.aclweb.org/anthology/E17-1039.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 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.5299999713897705,"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2741893366.pdf","grobid_xml":"https://content.openalex.org/works/W2741893366.grobid-xml"},"referenced_works_count":16,"referenced_works":["https://openalex.org/W207222","https://openalex.org/W4508078","https://openalex.org/W68436435","https://openalex.org/W191693080","https://openalex.org/W1565746575","https://openalex.org/W1881419322","https://openalex.org/W2039522160","https://openalex.org/W2096765155","https://openalex.org/W2103405567","https://openalex.org/W2108738385","https://openalex.org/W2124022436","https://openalex.org/W2130527179","https://openalex.org/W2165962877","https://openalex.org/W2962883620","https://openalex.org/W4234228486","https://openalex.org/W4290422161"],"related_works":["https://openalex.org/W4238897586","https://openalex.org/W435179959","https://openalex.org/W2619091065","https://openalex.org/W2059640416","https://openalex.org/W1490753184","https://openalex.org/W2284465472","https://openalex.org/W2291782699","https://openalex.org/W1993948687","https://openalex.org/W2000169967","https://openalex.org/W4385572368"],"abstract_inverted_index":{"Comparing":[0],"NLP":[1],"systems":[2,39],"to":[3,50,58,85,97,103,125,130,136],"select":[4],"the":[5,35,38,45,59,75,87,90,94,105,126,132],"best":[6,133],"one":[7],"for":[8,20],"a":[9,30,52,81],"task":[10],"of":[11,37,55,93],"interest,":[12],"such":[13],"as":[14],"named":[15],"entity":[16],"recognition,":[17],"is":[18],"critical":[19],"practitioners":[21],"and":[22,113],"researchers.":[23],"A":[24],"rigorous":[25],"approach":[26,48,119],"involves":[27],"setting":[28],"up":[29],"hypothesis":[31,46],"testing":[32,47],"scenario":[33],"using":[34],"performance":[36,91],"on":[40,74,110],"query":[41],"documents.":[42],"However,":[43],"often":[44],"needs":[49],"send":[51],"large":[53],"number":[54],"document":[56],"queries":[57,123],"systems,":[60,96],"which":[61],"can":[62],"be":[63,98],"problematic.":[64],"In":[65],"this":[66],"paper,":[67],"we":[68],"present":[69],"an":[70],"effective":[71],"alternative":[72],"based":[73],"multi-armed":[76],"bandit":[77],"(MAB).":[78],"We":[79],"propose":[80],"hierarchical":[82],"generative":[83],"model":[84],"represent":[86],"uncertainty":[88],"in":[89],"measures":[92],"competing":[95],"used":[99],"by":[100],"Thompson":[101],"Sampling":[102],"solve":[104],"resulting":[106],"MAB.":[107],"Experimental":[108],"results":[109],"both":[111],"synthetic":[112],"real":[114],"data":[115],"show":[116],"that":[117],"our":[118],"requires":[120],"significantly":[121],"fewer":[122],"compared":[124],"standard":[127],"benchmarking":[128],"technique":[129],"identify":[131],"system":[134],"according":[135],"Fmeasure.":[137]},"counts_by_year":[{"year":2019,"cited_by_count":1}],"updated_date":"2026-07-19T07:52:34.831488","created_date":"2025-10-10T00:00:00"}