{"id":"W4308571509","doi":"10.5334/johd.97","title":"CREMMA Medii Aevi: Literary Manuscript Text Recognition in Latin","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Latin Americans; Computer science; Representation (politics); Natural language processing; Artificial intelligence; Segmentation; Medieval Latin; Information retrieval; Linguistics; History; Classics; Political science","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.000636499,0.001102234,0.0004920916,0.00355832,0.0009871394,0.001898034,0.001177825,0.001120873,0.008566272],"category_scores_gemma":[0.002731885,0.0001681824,0.0008820532,0.002242932,0.0005243726,0.001411159,0.001271962,0.0007248707,0.01128483],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009279613,"about_ca_system_score_gemma":0.0009354121,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00735213,"about_ca_topic_score_gemma":0.01752676,"domain_scores_codex":[0.9993241,0.0001161166,0.00009722853,0.0002388841,0.000159885,0.00006382694],"domain_scores_gemma":[0.9988243,0.0002332793,0.0001065979,0.0003107163,0.0004113583,0.0001137551],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0008196988,0.000426958,0.01878261,0.002098014,0.0001613163,0.001457827,0.001321472,0.004214742,0.02934627,0.00385063,0.4893068,0.4482136],"study_design_scores_gemma":[0.0001303962,0.0002097609,0.04111538,0.0004584843,0.00008157677,0.002527266,0.002487289,0.02416502,0.03960425,0.002950195,0.8861455,0.0001247778],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.3447999,0.01377491,0.04615871,0.003938957,0.003984231,0.0009403466,0.4528052,0.05229851,0.08129923],"genre_scores_gemma":[0.2153117,0.001935884,0.07432689,0.0009791129,0.0006426053,0.0008428118,0.6760392,0.001317443,0.02860445],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008566272,"threshold_uncertainty_score":0.02865702,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.23807547872461,"score_gpt":0.3631402314952308,"score_spread":0.1250647527706208,"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."}}