{"id":"W4293580207","doi":"","title":"Textométrie et corpus générique","year":2013,"lang":"fr","type":"preprint","venue":"HAL (Le Centre pour la Communication Scientifique Directe)","topic":"Linguistics and Discourse Analysis","field":"Arts and Humanities","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Linguistic Association","funders":"","keywords":"Computer science; Trie; Programming language; Data structure","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":[0.002150389,0.001929652,0.001239355,0.009838399,0.002873886,0.00477928,0.001186479,0.001290296,0.07077909],"category_scores_gemma":[0.01501093,0.001075139,0.0009402041,0.01075218,0.001246577,0.002399928,0.001449028,0.002627112,0.0190928],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002639307,"about_ca_system_score_gemma":0.003394208,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.03081102,"about_ca_topic_score_gemma":0.01867391,"domain_scores_codex":[0.9947717,0.001560988,0.0004960825,0.001449006,0.001516673,0.000205588],"domain_scores_gemma":[0.991223,0.004657306,0.0003124131,0.001596089,0.002047871,0.0001633634],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.002475345,0.0004546162,0.007482904,0.003202181,0.0003314008,0.001037788,0.005675041,0.01257919,0.08134586,0.1031603,0.240751,0.5415044],"study_design_scores_gemma":[0.0005940363,0.000186669,0.0215603,0.0002624469,0.0001634401,0.00175117,0.002784778,0.02900882,0.0665478,0.01105094,0.8659189,0.0001706178],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"methods","genre_scores_codex":[0.207647,0.002400363,0.2993278,0.002994228,0.001260243,0.001421293,0.366833,0.03052066,0.08759543],"genre_scores_gemma":[0.254692,0.001423521,0.3187885,0.0002014325,0.000506336,0.005751915,0.3415739,0.01122137,0.06584107],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.07077909,"threshold_uncertainty_score":0.2367796,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02413780690869694,"score_gpt":0.248672250532369,"score_spread":0.224534443623672,"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."}}