{"id":"W3082920524","doi":"10.11606/d.45.2004.tde-20210729-142722","title":"Aprendizado de regras de substituição para normatização de textos históricos","year":2004,"lang":"pt","type":"dissertation","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Regina","funders":"","keywords":"Spelling; Lexicon; Linguistics; Computer science; Portuguese; Natural language processing; Computation; Artificial intelligence; Humanities; Art; Philosophy; Algorithm","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":[0.003443709,0.0007403959,0.000836048,0.002052317,0.001363524,0.002807637,0.001055266,0.0005816907,0.01146934],"category_scores_gemma":[0.01153182,0.0004295517,0.001131732,0.001888955,0.001683927,0.003746622,0.002000165,0.001366856,0.003604415],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009898777,"about_ca_system_score_gemma":0.001518869,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004454711,"about_ca_topic_score_gemma":0.008535743,"domain_scores_codex":[0.9979661,0.0006809603,0.0001818298,0.0004936962,0.0005048903,0.000172695],"domain_scores_gemma":[0.9924839,0.002216579,0.0005283362,0.002906657,0.001661784,0.0002025992],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001480188,0.0002895941,0.008012407,0.001224793,0.0001101653,0.000835093,0.008181303,0.003349564,0.06177686,0.07011279,0.01307007,0.8315571],"study_design_scores_gemma":[0.0003948862,0.001110592,0.0222774,0.0009087671,0.0006931777,0.002380566,0.01154799,0.07776324,0.2056637,0.1048072,0.572187,0.0002654599],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2558243,0.00224769,0.6419799,0.001632737,0.000799496,0.0009282616,0.002270872,0.02113383,0.07318299],"genre_scores_gemma":[0.484378,0.001270641,0.4850981,0.0002541648,0.0002112865,0.0002796801,0.002291175,0.0027549,0.02346204],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01146934,"threshold_uncertainty_score":0.03836876,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02037653076917156,"score_gpt":0.3108161817788661,"score_spread":0.2904396510096945,"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."}}