{"id":"W4390477219","doi":"10.33137/js.v5i.42259","title":"Corpus Linguistics Strategies for Identifying Accepted Theories in Early Modern England","year":2023,"lang":"en","type":"article","venue":"Scientonomy Journal for the Science of Science","topic":"Linguistic Variation and Morphology","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Corpus linguistics; Adjective; Context (archaeology); Set (abstract data type); Linguistics; Text corpus; Computational linguistics; Computer science; Natural language processing; Artificial intelligence; Noun phrase; Noun; History; Philosophy; Archaeology","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":["sts"],"consensus_categories":[],"category_scores_codex":[0.01589545,0.0005296014,0.0007496801,0.01511526,0.003656339,0.00911266,0.002333419,0.00148368,0.006652504],"category_scores_gemma":[0.06753214,0.0008819893,0.000447269,0.01955565,0.006941691,0.01065632,0.005818241,0.002261372,0.0009061672],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004940129,"about_ca_system_score_gemma":0.004397935,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01230078,"about_ca_topic_score_gemma":0.02193279,"domain_scores_codex":[0.9837883,0.01107933,0.001771917,0.001347972,0.00169907,0.000313415],"domain_scores_gemma":[0.9377862,0.05006621,0.003093629,0.00545455,0.00325736,0.000342149],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"observational","study_design_scores_codex":[0.0002852652,0.00009899446,0.0086995,0.001601746,0.00006833318,0.001206699,0.1824932,0.003413977,0.006290417,0.5862232,0.01424784,0.1953709],"study_design_scores_gemma":[0.0001298876,0.0001037593,0.01879428,0.002176078,0.0001022467,0.001583789,0.1370576,0.02357782,0.01030681,0.362137,0.4437541,0.0002766745],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2834665,0.008308627,0.6071526,0.01279875,0.0006334559,0.002345438,0.01274033,0.0008348344,0.07171942],"genre_scores_gemma":[0.5660443,0.002559678,0.4111437,0.0006027547,0.0002069335,0.003142139,0.007609363,0.0004468374,0.008244229],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9963437,"threshold_uncertainty_score":0.08406419,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07801143135558652,"score_gpt":0.3940014506527059,"score_spread":0.3159900192971194,"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."}}