{"id":"W4412609322","doi":"10.16995/dscn.18897","title":"Digital Text Analysis and Early Shakespeare Bibliography: Using Voyant Tools with Bad OCR","year":2025,"lang":"fr","type":"article","venue":"Digital Studies / Le champ numérique","topic":"Library Science and Information Systems","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"St. Francis Xavier University","funders":"","keywords":"Business","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"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":[0.008942234,0.0005155074,0.0004309189,0.02097414,0.00297125,0.01283233,0.000694458,0.0007460741,0.01663412],"category_scores_gemma":[0.07376501,0.0003990552,0.0003750667,0.01854957,0.003245521,0.01026863,0.004533559,0.001126916,0.005249857],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002812367,"about_ca_system_score_gemma":0.003179163,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008375058,"about_ca_topic_score_gemma":0.01129261,"domain_scores_codex":[0.9919158,0.002629039,0.001131449,0.0008363947,0.003260722,0.000226521],"domain_scores_gemma":[0.9503286,0.03017025,0.003745345,0.005782092,0.009466574,0.0005071385],"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":[0.0001493909,0.00004841005,0.009822923,0.001070829,0.00004168593,0.000394171,0.04110175,0.0005611916,0.005291658,0.05435314,0.03076881,0.8563961],"study_design_scores_gemma":[0.00004759203,0.0001169753,0.02551273,0.00283904,0.00008421504,0.0009741138,0.04307994,0.004858581,0.01562583,0.0505574,0.8560802,0.0002234035],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"other","genre_gemma":"methods","genre_scores_codex":[0.2675817,0.01430492,0.2687118,0.01478237,0.003378258,0.0009426124,0.01056659,0.02258116,0.3971506],"genre_scores_gemma":[0.5643985,0.005981474,0.3672282,0.0009078505,0.0007780431,0.0005898899,0.004898273,0.007318102,0.04789966],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.02097414,"threshold_uncertainty_score":0.05564672,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03526317689244406,"score_gpt":0.2607859084692743,"score_spread":0.2255227315768302,"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."}}