{"id":"W6910736779","doi":"10.48550/arxiv.2304.14044","title":"Large Scale Genealogical Information Extraction From Handwritten Quebec Parish Records","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Workflow; Consistency (knowledge bases); Information extraction; Sample (material); Scale (ratio); Process (computing)","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005740687,0.0009122546,0.0005638513,0.009626466,0.001626677,0.002437291,0.0008855264,0.0005627273,0.007323451],"category_scores_gemma":[0.003754219,0.0003289626,0.0005092077,0.00657583,0.0004749625,0.0006776061,0.0008076571,0.0004757688,0.004654624],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003599228,"about_ca_system_score_gemma":0.007433638,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.5414031,"about_ca_topic_score_gemma":0.6405631,"domain_scores_codex":[0.9991332,0.00006126092,0.00005911696,0.0002626181,0.0003486419,0.0001352871],"domain_scores_gemma":[0.9966733,0.0006831813,0.0002527302,0.0005452677,0.001673892,0.00017161],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003099271,0.0001217265,0.03796339,0.0008666253,0.0001263238,0.002187217,0.002954646,0.008017142,0.04603058,0.003321427,0.116189,0.781912],"study_design_scores_gemma":[0.00006517132,0.00008718109,0.2562456,0.0005904532,0.0001463072,0.001454484,0.0040362,0.1084579,0.08658358,0.004877649,0.5371774,0.0002780151],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2830127,0.002960307,0.3044915,0.001839195,0.0002658869,0.002197809,0.3001207,0.06302737,0.04208461],"genre_scores_gemma":[0.2765856,0.001227361,0.4131039,0.000232759,0.00009273638,0.0008065757,0.2620527,0.001690355,0.04420803],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.4585969,"threshold_uncertainty_score":0.9225954,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1030290525406084,"score_gpt":0.2252851522090067,"score_spread":0.1222560996683982,"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."}}