{"id":"W3190643984","doi":"10.5334/johd.42","title":"The Corpus for Idiolectal Research (CIDRE)","year":2021,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Agence Nationale de la Recherche","keywords":"Paratext; Scripting language; Metadata; License; Computer science; World Wide Web; Open source; Style (visual arts); Literature; Linguistics; Art; Programming language; Software; Philosophy","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":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.00602987,0.0000610319,0.0001449529,0.0000552555,0.00102209,0.002854078,0.007023896,0.00002985166,0.00005150675],"category_scores_gemma":[0.001127001,0.00004200391,0.00004596419,0.0002050661,0.00009743274,0.001758859,0.003619271,0.0003737221,0.00001343188],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005003015,"about_ca_system_score_gemma":0.0006641811,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001720555,"about_ca_topic_score_gemma":0.0001098324,"domain_scores_codex":[0.9984014,0.0003626245,0.0003585747,0.0001755522,0.0004431088,0.0002587294],"domain_scores_gemma":[0.9967007,0.001008727,0.0002039329,0.0009718923,0.001057059,0.00005765899],"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":[0.00006884465,0.00004797125,0.00007516629,0.00001626505,0.00007142233,0.0001126014,0.0009032951,0.000003570487,0.0003850298,0.7664054,0.2042795,0.02763091],"study_design_scores_gemma":[0.0003854038,0.0001459062,0.0001044278,0.00005091084,0.00000702195,0.0002474539,0.001144051,0.002178077,0.001680967,0.02973528,0.9642414,0.00007907762],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009849429,0.01502294,0.8970355,0.03802095,0.007333692,0.001276149,0.000461515,0.00005414722,0.03094565],"genre_scores_gemma":[0.734016,0.003578722,0.1930323,0.004324809,0.005703095,0.0000597439,0.0004585026,0.000098982,0.05872786],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7599619,"threshold_uncertainty_score":0.9983486,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6364291660620014,"score_gpt":0.5070253019086151,"score_spread":0.1294038641533863,"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."}}