{"id":"W3131420591","doi":"10.4000/corpus.5422","title":"Le corpus comme portail pour l’étude de la variation (socio)linguistique","year":2021,"lang":"fr","type":"article","venue":"Corpus","topic":"Linguistic Variation and Morphology","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Sociolinguistics; Vernacular; Linguistics; Annotation; Variation (astronomy); Variety (cybernetics); Computer science; Citation; Coding (social sciences); Corpus linguistics; Natural language processing; Sociology; Artificial intelligence; World Wide Web; Philosophy; Social science","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":[0.01306629,0.0008078245,0.001048905,0.009148172,0.005507349,0.009197074,0.001856338,0.00221388,0.01596224],"category_scores_gemma":[0.02828188,0.0005915429,0.0007193739,0.01477428,0.00645302,0.0060271,0.003632042,0.004534935,0.00472744],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003392391,"about_ca_system_score_gemma":0.006916443,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02160032,"about_ca_topic_score_gemma":0.01646946,"domain_scores_codex":[0.983595,0.009051781,0.001210762,0.001803627,0.004048041,0.000290847],"domain_scores_gemma":[0.9729311,0.01451895,0.0009814865,0.006619268,0.00430708,0.0006420726],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002265273,0.0000995815,0.004844396,0.001179202,0.00009037802,0.001173174,0.02309529,0.001139663,0.009904523,0.6649102,0.06428029,0.2290569],"study_design_scores_gemma":[0.00002423998,0.00004911295,0.005120806,0.0005344831,0.00002358567,0.001112814,0.004397302,0.0009843182,0.003138222,0.03518143,0.9493713,0.00006240681],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06205012,0.01822551,0.5585672,0.02649192,0.008808143,0.001978717,0.04051246,0.002342627,0.2810234],"genre_scores_gemma":[0.3015409,0.01366334,0.5458039,0.005506289,0.004899677,0.009342982,0.03681235,0.004421962,0.07800857],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02160032,"threshold_uncertainty_score":0.06910199,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03260862938634945,"score_gpt":0.3295949799492386,"score_spread":0.2969863505628891,"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."}}