{"id":"W4411377134","doi":"10.1038/s41597-025-05301-4","title":"The Austronesian and the Micronesian Comparative Dictionaries as CLDF datasets","year":2025,"lang":"en","type":"article","venue":"Scientific Data","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"Ministry of Education, India; Ministry of Education - Singapore","keywords":"Micronesian; Austronesian languages; Interoperability; Computer science; Comparative method; Linguistics; Comparative linguistics; Usability; Information retrieval; World Wide Web; History; Genealogy; Sociology; Philology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006250244,0.00008814784,0.0000847523,0.00001619272,0.001227502,0.0002904419,0.0007999209,0.00002767647,0.00000305262],"category_scores_gemma":[0.0000740945,0.00004911638,0.00001885066,0.0000989925,0.001583239,0.000001871927,0.001312214,0.00004964443,0.000009887714],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000003242913,"about_ca_system_score_gemma":0.00008272292,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004881163,"about_ca_topic_score_gemma":0.0004086789,"domain_scores_codex":[0.9991528,0.00006935142,0.0001228867,0.0004186315,0.0000800936,0.0001561729],"domain_scores_gemma":[0.9985393,0.00005326865,0.00004151468,0.001303667,0.00003707656,0.00002523398],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002353593,0.00003643384,0.0014993,0.00001114131,0.0003053792,7.566402e-7,0.0003488969,0.000009416773,0.04855326,0.02121127,0.9226094,0.005179414],"study_design_scores_gemma":[0.0003920103,0.00001296078,0.005297362,0.000003872785,0.00003319746,0.000003914814,0.0008610819,0.00008655206,0.003610535,0.001279457,0.9883537,0.00006533167],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8881877,0.04776976,0.00139275,0.02435471,0.00552183,0.00160724,0.01940037,0.00001538382,0.01175027],"genre_scores_gemma":[0.9906091,0.0003256625,0.0001519888,0.0001924047,0.00006976854,0.00001707552,0.00331493,0.000003582863,0.005315506],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1024214,"threshold_uncertainty_score":0.9441079,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0220802469138846,"score_gpt":0.3009350242765415,"score_spread":0.2788547773626569,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}