{"id":"W4318465586","doi":"10.1007/s10032-023-00427-w","title":"Large-scale genealogical information extraction from handwritten Quebec parish records","year":2023,"lang":"en","type":"article","venue":"International Journal on Document Analysis and Recognition (IJDAR)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université du Québec à Chicoutimi","funders":"Association Nationale de la Recherche et de la Technologie","keywords":"Computer science; Workflow; Consistency (knowledge bases); Scale (ratio); Sample (material); Information extraction; Population; Artificial intelligence; Natural language processing; Information retrieval; Data mining; Database; Geography; Medicine; Cartography","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.0001654454,0.0004757779,0.0003976939,0.0101882,0.001252563,0.001014416,0.0007713487,0.0005526625,0.004731523],"category_scores_gemma":[0.001251271,0.0002136699,0.000349915,0.009637579,0.0002928137,0.0004428312,0.0003527364,0.0004140715,0.002056866],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00286574,"about_ca_system_score_gemma":0.004257022,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.7610443,"about_ca_topic_score_gemma":0.8804678,"domain_scores_codex":[0.9997554,0.0000140221,0.00001663433,0.00006584187,0.00009088822,0.00005710652],"domain_scores_gemma":[0.9988509,0.0001992776,0.0001049193,0.0001242639,0.0006522605,0.000068351],"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":[0.0003811385,0.0002023995,0.1113929,0.0009723843,0.0002517002,0.003341616,0.001825512,0.01942278,0.04472473,0.002418816,0.09632257,0.7187436],"study_design_scores_gemma":[0.00005820305,0.0001213723,0.5587323,0.0003811162,0.0002994832,0.001559678,0.004329372,0.1796841,0.02781506,0.001816412,0.2250559,0.0001469979],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7282026,0.004598203,0.04237662,0.0008282549,0.0002290404,0.0006873042,0.190739,0.006828989,0.02551006],"genre_scores_gemma":[0.7460549,0.001416513,0.06518684,0.0001138294,0.0001071487,0.0002424705,0.160063,0.0003388126,0.02647656],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2389557,"threshold_uncertainty_score":0.4807258,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01208182651710639,"score_gpt":0.2916883980205446,"score_spread":0.2796065715034383,"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."}}