{"id":"W2736098086","doi":"","title":"Community-based corpus-building: Three case studies","year":2017,"lang":"en","type":"article","venue":"The COCOON platform (University of Paris)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Lemmatisation; Linguistics; Corpus linguistics; Annotation; Focus (optics); Ethos; Narrative; Documentation; Artificial intelligence; Natural language processing","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":[0.0241929,0.0007943053,0.0006263213,0.003900015,0.01608703,0.005442139,0.005823713,0.004498527,0.00627773],"category_scores_gemma":[0.0624693,0.0006885398,0.0007004514,0.006923052,0.007723842,0.006229883,0.01125919,0.003034838,0.001477635],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.007564634,"about_ca_system_score_gemma":0.009988043,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0712145,"about_ca_topic_score_gemma":0.14083,"domain_scores_codex":[0.9797395,0.01448117,0.0007965866,0.001233729,0.00239784,0.001351262],"domain_scores_gemma":[0.9230043,0.0495392,0.002266927,0.008576131,0.01112876,0.005484763],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"observational","study_design_scores_codex":[0.0004600095,0.00149901,0.02706688,0.001635695,0.00007156128,0.02505757,0.7013195,0.003471382,0.003561703,0.05163086,0.03065958,0.1535664],"study_design_scores_gemma":[0.0001995967,0.0002826125,0.01256114,0.001448642,0.00007245717,0.008028287,0.5738267,0.00586474,0.00564905,0.01670378,0.375172,0.0001908635],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8238559,0.002878253,0.07362028,0.01029876,0.000378935,0.005587359,0.002431456,0.0007151666,0.08023395],"genre_scores_gemma":[0.8721223,0.002023081,0.09897958,0.001398274,0.0001036398,0.003361438,0.002889119,0.0007448094,0.01837769],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.0712145,"threshold_uncertainty_score":0.1416,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07443160432197264,"score_gpt":0.3087848184024849,"score_spread":0.2343532140805122,"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."}}