{"id":"W2294508362","doi":"10.7202/1035932ar","title":"Proposition de protocole pour l’analyse des données textuelles : pour une démarche expérimentale en lexicométrie","year":2016,"lang":"fr","type":"article","venue":"Nouvelles perspectives en sciences sociales","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":28,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Humanities; Art","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","sts"],"consensus_categories":["sts"],"category_scores_codex":[0.003532526,0.000627127,0.0005697027,0.0004664513,0.001844794,0.0009277746,0.002654421,0.0004321697,0.0002265941],"category_scores_gemma":[0.001282375,0.0004623093,0.0003527023,0.002020819,0.00744803,0.002625801,0.0009252717,0.0004583759,0.0001345933],"about_ca_system_candidate":true,"about_ca_system_consensus":true,"about_ca_system_score_codex":0.003978719,"about_ca_system_score_gemma":0.006155431,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001944197,"about_ca_topic_score_gemma":0.0007000256,"domain_scores_codex":[0.993473,0.001992635,0.0006217837,0.001514042,0.001069592,0.001328958],"domain_scores_gemma":[0.9965411,0.00157054,0.000497521,0.0005184914,0.0005801101,0.0002922475],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00008811885,0.001353437,0.003912275,0.0003010415,0.0001829296,0.0001014036,0.1909269,0.00002155933,0.2280764,0.4121722,0.002877667,0.1599861],"study_design_scores_gemma":[0.001132016,0.0008009475,0.002598394,0.002118836,0.0001114687,0.0001914087,0.1394066,0.002046852,0.312556,0.5365885,0.001078881,0.001370099],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1855359,0.05696989,0.6511344,0.09293842,0.0004786603,0.005548447,0.00009744867,0.001332619,0.005964234],"genre_scores_gemma":[0.6105106,0.0008298142,0.3827467,0.0001484883,0.0006099847,0.0004082756,0.000001371375,0.00003586035,0.004708879],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.4249747,"threshold_uncertainty_score":0.9998448,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07846731784657346,"score_gpt":0.3714836107079668,"score_spread":0.2930162928613934,"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."}}