{"id":"https://openalex.org/W7224252831","doi":"https://doi.org/10.18653/2026.groundlm-1.4","title":"Trust-Aware Multi-Axis Reliability Auditing for NLP Classifiers: A Theoretical and Empirical Framework","display_name":"Trust-Aware Multi-Axis Reliability Auditing for NLP Classifiers: A Theoretical and Empirical Framework","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7224252831","doi":"https://doi.org/10.18653/2026.groundlm-1.4"},"language":"en","primary_location":{"id":"doi:10.18653/2026.groundlm-1.4","doi":"https://doi.org/10.18653/2026.groundlm-1.4","is_oa":true,"landing_page_url":"https://doi.org/10.18653/2026.groundlm-1.4","pdf_url":"https://aclanthology.org/2026.groundlm-1.4.pdf","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.groundlm-1.4.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5124910313","display_name":"Noor Islam S. Mohammad","orcid":"https://orcid.org/0009-0007-8823-6681","observed_orcids":["https://orcid.org/0009-0007-8823-6681"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Noor Islam S. Mohammad","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5124910313"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.88962101,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"43","last_page":"55"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.4790000021457672,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.4790000021457672,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.25049999356269836,"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/T10927","display_name":"Access Control and Trust","score":0.07109999656677246,"subfield":{"id":"https://openalex.org/subfields/3312","display_name":"Sociology and Political Science"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/classifier-reliability","display_name":"classifier reliability","score":0.6430000066757202},{"id":"https://openalex.org/keywords/trustworthiness","display_name":"trustworthiness","score":0.6230000257492065},{"id":"https://openalex.org/keywords/explainable-ai","display_name":"explainable AI","score":0.5899999737739563},{"id":"https://openalex.org/keywords/algorithmic-fairness","display_name":"algorithmic fairness","score":0.4000000059604645}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6486999988555908},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.6007000207901001},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5806000232696533},{"id":"https://openalex.org/C199521495","wikidata":"https://www.wikidata.org/wiki/Q181487","display_name":"Audit","level":2,"score":0.5163999795913696},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5109000205993652},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3458999991416931},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.33719998598098755},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.33480000495910645},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.30219998955726624},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2989000082015991},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.2630000114440918}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/2026.groundlm-1.4","doi":"https://doi.org/10.18653/2026.groundlm-1.4","is_oa":true,"landing_page_url":"https://doi.org/10.18653/2026.groundlm-1.4","pdf_url":"https://aclanthology.org/2026.groundlm-1.4.pdf","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/2026.groundlm-1.4","doi":"https://doi.org/10.18653/2026.groundlm-1.4","is_oa":true,"landing_page_url":"https://doi.org/10.18653/2026.groundlm-1.4","pdf_url":"https://aclanthology.org/2026.groundlm-1.4.pdf","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7224252831.pdf","grobid_xml":"https://content.openalex.org/works/W7224252831.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W3089264730","https://openalex.org/W7214775865","https://openalex.org/W6904135551","https://openalex.org/W6906331440","https://openalex.org/W7166846628","https://openalex.org/W7148592109","https://openalex.org/W7221338173","https://openalex.org/W7213601257","https://openalex.org/W7170284639","https://openalex.org/W6888553912"],"abstract_inverted_index":{"Accuracy":[0],"remains":[1],"the":[2,14,95,120,187],"dominant":[3],"criterion":[4,83],"for":[5,18,87,190],"selecting":[6],"NLP":[7,31,192],"classifiers,":[8],"yet":[9],"it":[10],"fails":[11],"to":[12,118,135],"characterize":[13],"operational":[15],"reliability":[16,38,73,184],"required":[17],"deployment":[19],"under":[20,150],"class":[21],"imbalance,":[22],"distribution":[23],"shift,":[24],"and":[25,54,57,94,113,132,144,162,181],"demographic":[26],"heterogeneity.We":[27],"introduce":[28],"TNMM++":[29,67],"(Trustworthy":[30],"Multi-axis":[32],"Monitoring,":[33],"extended),":[34],"a":[35,72,75,81,90],"theoretically":[36],"grounded":[37],"audit":[39],"that":[40,116,124,175],"evaluates":[41],"classifiers":[42],"along":[43],"seven":[44],"complementary":[45],"axes,":[46],"utility,":[47],"calibration,":[48],"robustness,":[49],"safety,":[50],"fairness,":[51],"selective":[52],"prediction,":[53],"uncertainty":[55],"quantification,":[56],"formalises":[58],"model":[59],"selection":[60,177],"as":[61,147,186],"constrained":[62],"multi-objective":[63],"optimization.Beyond":[64],"metric":[65],"aggregation,":[66],"contributes":[68],"six":[69],"algorithmic":[70],"results:":[71],"tensor,":[74],"Pareto":[76],"convex-hull":[77],"theorem,":[78],"varianceadaptive":[79],"weighting,":[80],"boundary-predictor":[82],"with":[84,98],"exclusion":[85],"rule":[86],"majority-class":[88],"collapse,":[89],"Bayesian":[91],"risk":[92],"decomposition,":[93],"TAMS":[96],"algorithm":[97],"optimality":[99],"bound.Evaluating":[100],"eight":[101],"lightweight":[102],"Transformer":[103],"encoders":[104],"(MiniLM,":[105],"ELECTRA,":[106],"mBERT,":[107],"DistilBERT,":[108],"MobileBERT,":[109],"ALBERT,":[110],"RoBERTa,":[111],"XLM-R)":[112],"excluding":[114],"models":[115,128],"failed":[117],"learn":[119],"task,":[121],"we":[122],"find":[123],"properly":[125],"trained":[126],"mid-tier":[127],"exhibit":[129],"balanced":[130],"utility":[131],"calibration":[133],"superior":[134],"collapsed":[136],"high-accuracy":[137],"counterparts.TrustScore":[138],"identifies":[139],"mBERT":[140],"(0.6685),":[141],"ELEC-TRA":[142],"(0.6630),":[143],"DistilBERT":[145],"(0.6627)":[146],"deployment-reliable.Sensitivity":[148],"analysis":[149],"1,000":[151],"Dirichlet":[152],"weight":[153],"perturbations":[154],"confirms":[155],"ranking":[156],"stability":[157],"in":[158],"92.4%":[159],"of":[160],"configurations,":[161],"multi-seed":[163],"evaluation":[164],"(5":[165],"seeds)":[166],"demonstrates":[167],"conclusions":[168],"are":[169],"not":[170],"seed":[171],"artifacts.Our":[172],"findings":[173],"establish":[174],"single-metric":[176],"is":[178],"statistically":[179],"unreliable":[180],"motivate":[182],"multi-axis":[183],"auditing":[185],"default":[188],"protocol":[189],"trustworthy":[191],"deployment.":[193]},"counts_by_year":[],"updated_date":"2026-11-01T11:40:01.658920","created_date":"2026-10-30T00:00:00"}