{"id":"https://openalex.org/W4416036701","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.418","title":"A Rigorous Evaluation of LLM Data Generation Strategies for Low-Resource Languages","display_name":"A Rigorous Evaluation of LLM Data Generation Strategies for Low-Resource Languages","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416036701","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.418"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.emnlp-main.418","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.418","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.418","pdf_url":"https://aclanthology.org/2025.emnlp-main.418.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2025.emnlp-main.418.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5036049138","display_name":"Tatiana Anikina","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I33256026","display_name":"German Research Centre for Artificial Intelligence","ror":"https://ror.org/01ayc5b57","country_code":"DE","type":"funder","lineage":["https://openalex.org/I33256026"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Tatiana Anikina","raw_affiliation_strings":["German Research Institute for Artificial Intelligence (DFKI) , Saarbr\u00fccken , Germany"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"German Research Institute for Artificial Intelligence (DFKI) , Saarbr\u00fccken , Germany","institution_ids":["https://openalex.org/I33256026"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081841143","display_name":"J\u00e1n \u010cegi\u0148","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4401200340","display_name":"Kempelen Institute of Intelligent Technologies","ror":"https://ror.org/02j78kr76","country_code":"SK","type":"nonprofit","lineage":["https://openalex.org/I4401200340"]},{"id":"https://openalex.org/I60587646","display_name":"Brno University of Technology","ror":"https://ror.org/03613d656","country_code":"CZ","type":"education","lineage":["https://openalex.org/I60587646"]}],"countries":["CZ","SK"],"is_corresponding":false,"raw_author_name":"Jan Cegin","raw_affiliation_strings":["Faculty of Information Technology , Brno University of Technology , Brno , Czechia","Kempelen Institute of Intelligent Technologies , Bratislava , Slovakia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Information Technology , Brno University of Technology , Brno , Czechia","institution_ids":["https://openalex.org/I60587646"]},{"raw_affiliation_string":"Kempelen Institute of Intelligent Technologies , Bratislava , Slovakia","institution_ids":["https://openalex.org/I4401200340"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043199710","display_name":"Jakub \u0160imko","orcid":"https://orcid.org/0000-0003-0239-4237","observed_orcids":["https://orcid.org/0000-0003-0239-4237"]},"institutions":[{"id":"https://openalex.org/I4401200340","display_name":"Kempelen Institute of Intelligent Technologies","ror":"https://ror.org/02j78kr76","country_code":"SK","type":"nonprofit","lineage":["https://openalex.org/I4401200340"]}],"countries":["SK"],"is_corresponding":false,"raw_author_name":"Jakub Simko","raw_affiliation_strings":["Kempelen Institute of Intelligent Technologies , Bratislava , Slovakia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Kempelen Institute of Intelligent Technologies , Bratislava , Slovakia","institution_ids":["https://openalex.org/I4401200340"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5102820995","display_name":"Simon Ostermann","orcid":"https://orcid.org/0000-0002-0899-0657","observed_orcids":["https://orcid.org/0000-0002-0899-0657"]},"institutions":[{"id":"https://openalex.org/I33256026","display_name":"German Research Centre for Artificial Intelligence","ror":"https://ror.org/01ayc5b57","country_code":"DE","type":"funder","lineage":["https://openalex.org/I33256026"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Simon Ostermann","raw_affiliation_strings":["German Research Institute for Artificial Intelligence (DFKI) , Saarbr\u00fccken , Germany"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"German Research Institute for Artificial Intelligence (DFKI) , Saarbr\u00fccken , Germany","institution_ids":["https://openalex.org/I33256026"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.9066,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.87596846,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":96,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"8293","last_page":"8314"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.7167999744415283,"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":0.7167999744415283,"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":0.2736999988555908,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.002300000051036477,"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/synthetic-data-generation","display_name":"synthetic data generation","score":0.8629999756813049},{"id":"https://openalex.org/keywords/low-resource-languages","display_name":"low-resource languages","score":0.9390000104904175},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8600000143051147},{"id":"https://openalex.org/keywords/data-augmentation","display_name":"data augmentation","score":0.5379999876022339},{"id":"https://openalex.org/keywords/prompting-strategies","display_name":"prompting strategies","score":0.6470000147819519},{"id":"https://openalex.org/keywords/demonstrations","display_name":"demonstrations","score":0.527999997138977},{"id":"https://openalex.org/keywords/self-revision","display_name":"self-revision","score":0.9819999933242798},{"id":"https://openalex.org/keywords/typologically-diverse-languages","display_name":"typologically diverse languages","score":0.7490000128746033}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6308000087738037},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3513000011444092},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.30869999527931213},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.289900004863739},{"id":"https://openalex.org/C133462117","wikidata":"https://www.wikidata.org/wiki/Q4929239","display_name":"Data collection","level":2,"score":0.28839999437332153},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.27959999442100525},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2694999873638153},{"id":"https://openalex.org/C2776187449","wikidata":"https://www.wikidata.org/wiki/Q1513879","display_name":"Natural language generation","level":3,"score":0.2676999866962433},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.25270000100135803},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.24140000343322754}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.18653/v1/2025.emnlp-main.418","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.418","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.418","pdf_url":"https://aclanthology.org/2025.emnlp-main.418.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},{"id":"pmh:oai:zenodo.org:21234228","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/arXiv:2506.12158","pdf_url":"https://arxiv.org/pdf/2506.12158","source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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":"EMNLP 2025, Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, Suzhou, China, 4-9 November, 2025","raw_type":"info:eu-repo/semantics/conferencePaper"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.emnlp-main.418","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.418","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.418","pdf_url":"https://aclanthology.org/2025.emnlp-main.418.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G3931493262","display_name":"Improving scientific excellence and creativity in combating disinformation with artificial intelligence and language technologies","funder_award_id":"101079164","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"},{"id":"https://openalex.org/G7381952728","display_name":"AI-CODE - AI services for COntinuous trust in emerging Digital Environments","funder_award_id":"101135437","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"},{"id":"https://openalex.org/G941802975","display_name":null,"funder_award_id":"90254","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"}],"funders":[{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320334322","display_name":"HORIZON EUROPE Framework Programme","ror":null}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416036701.pdf","grobid_xml":"https://content.openalex.org/works/W4416036701.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7140130829","https://openalex.org/W7165198664","https://openalex.org/W7160101653","https://openalex.org/W4402671762","https://openalex.org/W7215011477","https://openalex.org/W4411360037","https://openalex.org/W4417298833","https://openalex.org/W4403593504","https://openalex.org/W7128432232","https://openalex.org/W6893319183"],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"are":[4],"increasingly":[5],"used":[6],"to":[7,111],"generate":[8],"synthetic":[9,137],"textual":[10],"data":[11,110,138],"for":[12,23,46,136],"training":[13],"smaller":[14,144],"specialized":[15],"models.However,":[16],"a":[17],"comparison":[18],"of":[19,56,95,129],"various":[20,29],"generation":[21,58,96,134,139],"strategies":[22,31,59,135],"low-resource":[24,47,71],"language":[25],"settings":[26],"is":[27],"lacking.While":[28],"prompting":[30,123],"have":[32],"been":[33],"proposed-such":[34],"as":[35,112,114],"demonstrations,":[36],"labelbased":[37],"summaries,":[38],"and":[39,60,76],"self-revision-their":[40],"comparative":[41],"effectiveness":[42],"remains":[43],"unclear,":[44],"especially":[45],"languages.In":[48],"this":[49],"paper,":[50],"we":[51,80],"systematically":[52],"evaluate":[53],"the":[54,106,127],"performance":[55,84],"these":[57],"their":[61],"combinations":[62,94],"across":[63],"11":[64],"typologically":[65],"diverse":[66],"languages,":[67],"including":[68],"several":[69],"extremely":[70],"ones.Using":[72],"three":[73],"NLP":[74],"tasks":[75],"four":[77],"open-source":[78],"LLMs,":[79,131],"assess":[81],"downstream":[82],"model":[83],"on":[85],"generated":[86],"versus":[87],"gold-standard":[88],"data.Our":[89],"results":[90],"show":[91],"that":[92,121],"strategic":[93],"methods-particularly":[97],"targetlanguage":[98],"demonstrations":[99],"with":[100,108,143],"LLM-based":[101],"revisions-yield":[102],"strong":[103],"performance,":[104],"narrowing":[105],"gap":[107],"real":[109],"little":[113],"5%":[115],"in":[116,140],"some":[117],"settings.We":[118],"also":[119],"find":[120],"smart":[122],"techniques":[124],"can":[125],"reduce":[126],"advantage":[128],"larger":[130],"highlighting":[132],"efficient":[133],"lowresource":[141],"scenarios":[142],"models.":[145]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2025-11-08T00:00:00"}