{"id":"W2207671015","doi":"10.1055/s-0038-1638748","title":"Key Concepts to Assess the Readiness of Data for International Research: Data Quality, Lineage and Provenance, Extraction and Processing Errors, Traceability, and Curation","year":2011,"lang":"en","type":"article","venue":"Yearbook of Medical Informatics","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":48,"is_retracted":false,"has_abstract":true,"ca_institutions":"Algoma University; Esri (Canada)","funders":"Directorate-General for Information Society and Media; European Commission","keywords":"Metadata; Computer science; Data quality; Data science; Data curation; Metadata repository; Data warehouse; Traceability; Data extraction; Information retrieval; Data element; Database; World Wide Web; Engineering","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"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.4818004,0.0017987,0.003288372,0.03143398,0.006854969,0.0307845,0.006188692,0.006589013,0.005839377],"category_scores_gemma":[0.6337735,0.001837364,0.003605607,0.03235938,0.03496922,0.04481247,0.0180998,0.007040319,0.001823895],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.01894035,"about_ca_system_score_gemma":0.05730015,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006962045,"about_ca_topic_score_gemma":0.004837114,"domain_scores_codex":[0.4681175,0.3375718,0.1098284,0.01524121,0.06477115,0.004469908],"domain_scores_gemma":[0.1371502,0.6358627,0.08860188,0.04456062,0.08931543,0.004509134],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0004960846,0.0003289135,0.06694052,0.05103077,0.00101944,0.0006227718,0.07417209,0.00359752,0.002802388,0.4557418,0.02154761,0.3217001],"study_design_scores_gemma":[0.0001740742,0.0006600526,0.04221663,0.09892666,0.001124523,0.001627415,0.09489612,0.004490057,0.007525766,0.4475339,0.3000693,0.0007555101],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0466547,0.04965177,0.6776959,0.1394796,0.002618447,0.02105469,0.007621465,0.001181284,0.05404218],"genre_scores_gemma":[0.2220693,0.01365433,0.7378473,0.007128619,0.0006161967,0.01303553,0.003744241,0.0003756099,0.001528819],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.5181996,"threshold_uncertainty_score":0.6390324,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.8142543360737813,"score_gpt":0.6175428778911553,"score_spread":0.196711458182626,"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."}}