{"id":"W4311843543","doi":"","title":"Vers une étude comparative de différentes approches de classification automatique de textes provenant des secteurs métiers","year":2021,"lang":"fr","type":"preprint","venue":"HAL (Le Centre pour la Communication Scientifique Directe)","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Berger (Canada)","funders":"","keywords":"Computer 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":[0.01407195,0.001740689,0.001556521,0.01148688,0.001756884,0.009192187,0.002061081,0.002407896,0.004934371],"category_scores_gemma":[0.04031797,0.0006948373,0.002787846,0.007520422,0.001251968,0.006147446,0.001083321,0.002139328,0.00350981],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002239941,"about_ca_system_score_gemma":0.001906301,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02317349,"about_ca_topic_score_gemma":0.02676927,"domain_scores_codex":[0.9736268,0.00918506,0.002461589,0.00477827,0.009205276,0.0007429564],"domain_scores_gemma":[0.9393572,0.04206103,0.002056353,0.005007391,0.01104027,0.0004777938],"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":[0.00148888,0.0004614719,0.02759572,0.001425805,0.001665861,0.0001457006,0.001691018,0.01525878,0.0311836,0.004275773,0.005241942,0.9095654],"study_design_scores_gemma":[0.000497681,0.002820586,0.1983427,0.00149442,0.003973312,0.003127844,0.005850516,0.5084921,0.1505247,0.01240018,0.1118871,0.0005888813],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4196376,0.02389117,0.5178711,0.002272892,0.001063602,0.0008172159,0.005796607,0.01184546,0.01680439],"genre_scores_gemma":[0.5168118,0.005110197,0.4519857,0.0005269052,0.0005328316,0.0004594947,0.007769825,0.001200939,0.0156022],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02317349,"threshold_uncertainty_score":0.07442051,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03285696509734804,"score_gpt":0.2624052596048865,"score_spread":0.2295482945075384,"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."}}