{"id":"https://openalex.org/W4412889174","doi":"https://doi.org/10.18653/v1/2025.bea-1.30","title":"Automated L2 Proficiency Scoring: Weak Supervision, Large Language Models, and Statistical Guarantees","display_name":"Automated L2 Proficiency Scoring: Weak Supervision, Large Language Models, and Statistical Guarantees","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412889174","doi":"https://doi.org/10.18653/v1/2025.bea-1.30"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.bea-1.30","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.bea-1.30","pdf_url":"https://aclanthology.org/2025.bea-1.30.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 20th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2025)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.bea-1.30.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5078761239","display_name":"Aitor Arronte Alvarez","orcid":"https://orcid.org/0000-0002-1605-1772"},"institutions":[{"id":"https://openalex.org/I117965899","display_name":"University of Hawai\u02bbi at M\u0101noa","ror":"https://ror.org/01wspgy28","country_code":"US","type":"education","lineage":["https://openalex.org/I117965899"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Aitor Arronte Alvarez","raw_affiliation_strings":["University of Hawai'i at Mnoa Honolulu , HI , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Hawai'i at Mnoa Honolulu , HI , USA","institution_ids":["https://openalex.org/I117965899"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5052591149","display_name":"Naiyi Xie Fincham","orcid":"https://orcid.org/0000-0002-4959-436X"},"institutions":[{"id":"https://openalex.org/I117965899","display_name":"University of Hawai\u02bbi at M\u0101noa","ror":"https://ror.org/01wspgy28","country_code":"US","type":"education","lineage":["https://openalex.org/I117965899"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Naiyi Xie Fincham","raw_affiliation_strings":["University of Hawai'i at Mnoa Honolulu , HI , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Hawai'i at Mnoa Honolulu , HI , USA","institution_ids":["https://openalex.org/I117965899"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I117965899"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.14556707,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"384","last_page":"397"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9882000088691711,"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":0.9882000088691711,"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/computer-science","display_name":"Computer science","score":0.741554856300354},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.5471396446228027},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4469816982746124},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.350716233253479}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.741554856300354},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5471396446228027},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4469816982746124},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.350716233253479}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.bea-1.30","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.bea-1.30","pdf_url":"https://aclanthology.org/2025.bea-1.30.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 20th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2025)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.bea-1.30","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.bea-1.30","pdf_url":"https://aclanthology.org/2025.bea-1.30.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 20th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2025)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Quality Education","score":0.5199999809265137,"id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412889174.pdf","grobid_xml":"https://content.openalex.org/works/W4412889174.grobid-xml"},"referenced_works_count":1,"referenced_works":["https://openalex.org/W2959716049"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Weakly":[0],"supervised":[1],"learning":[2,7,82,115],"(WSL)":[3],"is":[4,13,45,133,164,205],"a":[5,108,113,119,126,151],"machine":[6,114],"approach":[8],"used":[9],"when":[10],"labeled":[11,203],"data":[12,204],"scarce":[14],"or":[15],"expensive":[16],"to":[17,88,135,189],"obtain.In":[18],"such":[19],"scenarios,":[20],"models":[21,104],"are":[22,34],"trained":[23,117],"using":[24,101],"weaker":[25],"supervision":[26],"sources":[27,33],"instead":[28],"of":[29,52,74,144,153,171],"human-annotated":[30],"data.However,":[31],"these":[32],"often":[35],"noisy":[36],"and":[37,61,80,106],"may":[38],"introduce":[39],"unquantified":[40],"biases":[41],"during":[42],"training.This":[43],"issue":[44],"particularly":[46],"pronounced":[47],"in":[48,111,198],"automated":[49],"scoring":[50,73,100],"(AS)":[51],"second":[53],"language":[54,103],"(L2)":[55],"learner":[56,76,155],"output,":[57],"where":[58,202],"high":[59],"variability":[60],"limited":[62],"generalizability":[63],"pose":[64],"significant":[65],"challenges.In":[66],"this":[67],"paper,":[68],"we":[69,175],"investigate":[70],"the":[71,142,145,169,181,190],"analytical":[72],"L2":[75,154],"responses":[77],"under":[78],"weak":[79],"semi-supervised":[81,109],"conditions,":[83],"leveraging":[84],"Prediction-Powered":[85],"Inference":[86],"(PPI)":[87],"provide":[89],"statistical":[90],"guarantees":[91],"on":[92,118,150],"score":[93],"validity.We":[94],"compare":[95],"two":[96],"approaches:":[97],"(1)":[98],"synthetic":[99],"large":[102],"(LLMs),":[105],"(2)":[107],"setting":[110],"which":[112],"model,":[116],"small":[120],"gold-standard":[121],"set,":[122],"generates":[123],"predictions":[124],"for":[125,140,167],"larger":[127],"unlabeled":[128],"corpus.In":[129],"both":[130],"cases,":[131],"PPI":[132,163,178],"applied":[134],"construct":[136],"valid":[137],"confidence":[138],"intervals":[139],"assessing":[141],"reliability":[143],"predicted":[146],"scores.Our":[147],"analysis,":[148],"based":[149],"dataset":[152],"conversations":[156],"with":[157],"an":[158],"AI":[159],"agent,":[160],"shows":[161],"that":[162,177],"highly":[165],"informative":[166],"evaluating":[168],"quality":[170],"weakly":[172],"annotated":[173],"data.Moreover,":[174],"demonstrate":[176],"can":[179],"increase":[180],"effective":[182],"sample":[183],"size":[184],"by":[185],"over":[186],"150%":[187],"relative":[188],"original":[191],"human-scored":[192],"subset,":[193],"enabling":[194],"more":[195],"robust":[196],"inference":[197],"educational":[199],"assessment":[200],"settings":[201],"scarce.":[206]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}