{"id":"W4414769236","doi":"10.31274/jlsc.19779","title":"Evaluating Multilingual Metadata Quality in Crossref","year":2025,"lang":"en","type":"article","venue":"Journal of Librarianship and Scholarly Communication","topic":"Intellectual Property Law","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University; Trinity College","funders":"","keywords":"Metadata; Multilingualism; Categorization; Publishing; Sample (material); Quality (philosophy)","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":["metaresearch","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.1515312,0.0004746056,0.0008540347,0.02135019,0.0030191,0.01011069,0.002010703,0.001229466,0.005046879],"category_scores_gemma":[0.4402376,0.0005683294,0.0009885408,0.02226445,0.003745179,0.01518968,0.0119598,0.001049496,0.001227186],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003229177,"about_ca_system_score_gemma":0.005309946,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005862376,"about_ca_topic_score_gemma":0.005264562,"domain_scores_codex":[0.8634559,0.05508566,0.02903403,0.008093589,0.04231546,0.0020154],"domain_scores_gemma":[0.3803115,0.3244323,0.1127075,0.05400806,0.1241729,0.004367613],"domain_codex":null,"domain_gemma":"reporting","domain_candidate":"reporting","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0004148252,0.0002103573,0.8903397,0.000919928,0.0002410372,0.0001712112,0.01717421,0.0008839886,0.001193383,0.006343473,0.001660667,0.08044733],"study_design_scores_gemma":[0.0001283804,0.001031105,0.8226351,0.003032886,0.0005676646,0.001647491,0.05101817,0.02346376,0.0177954,0.02255239,0.05575849,0.0003690973],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9123515,0.001587391,0.05842046,0.001934649,0.0001118144,0.001194451,0.004824146,0.0006118483,0.01896374],"genre_scores_gemma":[0.9568946,0.0003783047,0.03665908,0.0002612266,0.00008885278,0.000657231,0.003915533,0.0001464675,0.0009987447],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9898893,"threshold_uncertainty_score":0.8013834,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2579584788334103,"score_gpt":0.4846089632726107,"score_spread":0.2266504844392003,"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."}}