{"id":"https://openalex.org/W7204204986","doi":"https://doi.org/10.1109/ines69513.2026.11661205","title":"Optimizing Training Data Diversity with DataSpinner: An LLM-driven Curator-Agent Approach","display_name":"Optimizing Training Data Diversity with DataSpinner: An LLM-driven Curator-Agent Approach","publication_year":2026,"publication_date":"2026-07-01","ids":{"openalex":"https://openalex.org/W7204204986","doi":"https://doi.org/10.1109/ines69513.2026.11661205"},"language":"en","primary_location":{"id":"doi:10.1109/ines69513.2026.11661205","doi":"https://doi.org/10.1109/ines69513.2026.11661205","is_oa":true,"landing_page_url":"https://doi.org/10.1109/ines69513.2026.11661205","pdf_url":null,"source":{"id":"https://openalex.org/S7407087665","display_name":"IEEE International Conference on Intelligent Engineering Systems (INES)","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":"2026 IEEE 30th International Conference on Intelligent Engineering Systems (INES)","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","is_traditional":true,"indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://doi.org/10.1109/ines69513.2026.11661205","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5119542083","display_name":"Daniel Horti","orcid":"https://orcid.org/0009-0001-0354-7162","observed_orcids":["https://orcid.org/0009-0001-0354-7162"]},"institutions":[{"id":"https://openalex.org/I103356709","display_name":"Obuda University","ror":"https://ror.org/00ax71d21","country_code":"HU","type":"education","lineage":["https://openalex.org/I103356709"]}],"countries":["HU"],"is_corresponding":false,"raw_author_name":"D\u00e1niel Horti","raw_affiliation_strings":["University Research and Innovation Center Obuda University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University Research and Innovation Center Obuda University","institution_ids":["https://openalex.org/I103356709"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5117596554","display_name":"Kristi\u00e1n Sopkovi\u010d","orcid":"https://orcid.org/0009-0007-0835-3491","observed_orcids":["https://orcid.org/0009-0007-0835-3491"]},"institutions":[{"id":"https://openalex.org/I183764125","display_name":"Technical University of Ko\u0161ice","ror":"https://ror.org/05xm08015","country_code":"SK","type":"education","lineage":["https://openalex.org/I183764125"]}],"countries":["SK"],"is_corresponding":false,"raw_author_name":"Kristi\u00e1n Sopkovi\u010d","raw_affiliation_strings":["Technical University of Ko\u0161ice"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Technical University of Ko\u0161ice","institution_ids":["https://openalex.org/I183764125"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5080567273","display_name":"Daniel Hl\u00e1dek","orcid":"https://orcid.org/0000-0003-1148-3194","observed_orcids":["https://orcid.org/0000-0003-1148-3194"]},"institutions":[{"id":"https://openalex.org/I183764125","display_name":"Technical University of Ko\u0161ice","ror":"https://ror.org/05xm08015","country_code":"SK","type":"education","lineage":["https://openalex.org/I183764125"]}],"countries":["SK"],"is_corresponding":false,"raw_author_name":"Daniel Hl\u00e1dek","raw_affiliation_strings":["Technical University of Ko\u0161ice"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Technical University of Ko\u0161ice","institution_ids":["https://openalex.org/I183764125"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5056828566","display_name":"Artur Istvan Karoly","orcid":"https://orcid.org/0000-0002-2902-7253","observed_orcids":["https://orcid.org/0000-0002-2902-7253"]},"institutions":[{"id":"https://openalex.org/I103356709","display_name":"Obuda University","ror":"https://ror.org/00ax71d21","country_code":"HU","type":"education","lineage":["https://openalex.org/I103356709"]}],"countries":["HU"],"is_corresponding":false,"raw_author_name":"Art\u00far I. K\u00e1roly","raw_affiliation_strings":["University Research and Innovation Center Obuda University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University Research and Innovation Center Obuda University","institution_ids":["https://openalex.org/I103356709"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.0,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"483","last_page":"490"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9175000190734863,"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":0.9175000190734863,"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":0.019099999219179153,"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":0.013799999840557575,"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/training-data-diversity","display_name":"training data diversity","score":0.7760000228881836},{"id":"https://openalex.org/keywords/synthetic-data-generation","display_name":"synthetic data generation","score":0.7400000095367432},{"id":"https://openalex.org/keywords/data-augmentation","display_name":"data augmentation","score":0.6620000004768372},{"id":"https://openalex.org/keywords/model-size-effects","display_name":"model size effects","score":0.49300000071525574}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5997999906539917},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.5608999729156494},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4510999917984009},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.38089999556541443},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.35100001096725464},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3382999897003174},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.32989999651908875},{"id":"https://openalex.org/C133462117","wikidata":"https://www.wikidata.org/wiki/Q4929239","display_name":"Data collection","level":2,"score":0.27869999408721924},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.2685000002384186},{"id":"https://openalex.org/C2781316041","wikidata":"https://www.wikidata.org/wiki/Q1230584","display_name":"Diversity (politics)","level":2,"score":0.26750001311302185},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.2603999972343445},{"id":"https://openalex.org/C24756922","wikidata":"https://www.wikidata.org/wiki/Q1757694","display_name":"Data quality","level":3,"score":0.2549000084400177}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/ines69513.2026.11661205","doi":"https://doi.org/10.1109/ines69513.2026.11661205","is_oa":true,"landing_page_url":"https://doi.org/10.1109/ines69513.2026.11661205","pdf_url":null,"source":{"id":"https://openalex.org/S7407087665","display_name":"IEEE International Conference on Intelligent Engineering Systems (INES)","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":"2026 IEEE 30th International Conference on Intelligent Engineering Systems (INES)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1109/ines69513.2026.11661205","doi":"https://doi.org/10.1109/ines69513.2026.11661205","is_oa":true,"landing_page_url":"https://doi.org/10.1109/ines69513.2026.11661205","pdf_url":null,"source":{"id":"https://openalex.org/S7407087665","display_name":"IEEE International Conference on Intelligent Engineering Systems (INES)","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":"2026 IEEE 30th International Conference on Intelligent Engineering Systems (INES)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":15,"referenced_works":["https://openalex.org/W1978394996","https://openalex.org/W2087388117","https://openalex.org/W2284660317","https://openalex.org/W2296073425","https://openalex.org/W2963456134","https://openalex.org/W2963545917","https://openalex.org/W2963903950","https://openalex.org/W2970641574","https://openalex.org/W2971296908","https://openalex.org/W4385572634","https://openalex.org/W4402670307","https://openalex.org/W4412444947","https://openalex.org/W4412944735","https://openalex.org/W7133208539","https://openalex.org/W7133233472"],"related_works":["https://openalex.org/W4415798816","https://openalex.org/W7204110536","https://openalex.org/W7085511945","https://openalex.org/W6966201543","https://openalex.org/W4417447676","https://openalex.org/W7196937272","https://openalex.org/W4415800565","https://openalex.org/W7197003104","https://openalex.org/W46298369","https://openalex.org/W4416034256"],"abstract_inverted_index":{"Synthetic":[0],"data":[1,113,149],"generation":[2,19],"has":[3],"emerged":[4],"as":[5,57,83],"a":[6,30,43,58,84,107,124,132],"scalable":[7],"alternative":[8],"to":[9,69,111],"human-curated":[10],"datasets":[11,175],"for":[12,75],"fine-tuning":[13,164],"task-specific":[14],"language":[15],"models.":[16],"However,":[17],"uncontrolled":[18],"pipelines":[20],"risk":[21],"distributional":[22],"collapse,":[23],"where":[24],"the":[25,50,64,67,89,121,141],"training":[26,148,182],"corpus":[27,91],"converges":[28],"toward":[29],"narrow,":[31],"low-variance":[32],"distribution":[33,92],"that":[34,53,114,140,162],"undermines":[35],"model":[36,158],"generalization.":[37],"In":[38],"this":[39],"paper,":[40],"we":[41],"propose":[42],"curator":[44,79],"agent,":[45],"an":[46],"LLM-driven":[47],"component":[48],"of":[49,135],"DataSpinner":[51],"framework":[52,122],"treats":[54],"dataset":[55],"diversity":[56],"first-class":[59],"optimization":[60],"objective,":[61],"and":[62,103,105,128],"evaluate":[63],"approach":[65],"on":[66],"NLIs":[68],"JSON-formatted":[70],"action":[71],"sequence":[72],"compilation":[73],"task":[74],"mobile":[76],"robots.":[77],"The":[78],"agent":[80,110],"is":[81],"designed":[82],"closed-loop":[85],"framework:":[86],"it":[87,130],"analyzes":[88],"current":[90],"across":[93,155],"five":[94,156],"interpretable":[95],"dimensions":[96],"(complexity,":[97],"logic":[98],"type,":[99],"domain,":[100],"sentence":[101],"structure,":[102],"specificity)":[104],"steers":[106],"secondary":[108],"augmentation":[109],"generate":[112],"fills":[115],"identified":[116],"coverage":[117],"gaps.":[118],"We":[119],"instantiate":[120],"in":[123],"single-iteration":[125],"A/B":[126],"configuration":[127],"compare":[129],"against":[131],"non-curated":[133],"baseline":[134],"equal":[136],"size.":[137],"Results":[138],"show":[139],"curator-driven":[142],"pipeline":[143],"produces":[144],"measurably":[145],"more":[146],"diverse":[147],"than":[150],"classical":[151],"augmentation,":[152],"while":[153,176],"evaluation":[154],"Qwen2.5Instruct":[157],"sizes":[159],"(0.5B\u201314B)":[160],"revealed":[161],"although":[163],"improves":[165],"performance":[166],"over":[167],"zero-shot":[168],"prompting,":[169],"smaller":[170],"models":[171,178],"favor":[172],"tightly":[173],"focused":[174],"larger":[177],"can":[179],"accommodate":[180],"greater":[181],"diversity.":[183]},"counts_by_year":[],"updated_date":"2026-11-10T11:39:43.021295","created_date":"2026-08-26T00:00:00"}