{"id":"W3197767598","doi":"10.2196/31750","title":"Approaches and Criteria for Provenance in Biomedical Data Sets and Workflows: Protocol for a Scoping Review","year":2021,"lang":"en","type":"review","venue":"JMIR Research Protocols","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Bundesministerium für Bildung und Forschung; Deutsche Forschungsgemeinschaft","keywords":"Computer science; Workflow; Metadata; Data science; Data quality; Traceability; Transparency (behavior); Information retrieval; World Wide Web; Software engineering; Database; Engineering","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","metaepi_narrow","scholarly_communication"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.05718867,0.0004193396,0.002213208,0.0008069561,0.0003722109,0.002446082,0.003592677,0.0002409415,0.00006189154],"category_scores_gemma":[0.02043201,0.0002783134,0.0001728252,0.002549007,0.0005061833,0.0005335908,0.006375412,0.0005814061,0.00001230293],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001016222,"about_ca_system_score_gemma":0.001470484,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000003985477,"about_ca_topic_score_gemma":0.00001715474,"domain_scores_codex":[0.9891945,0.002091828,0.002128219,0.003384335,0.002221534,0.0009795844],"domain_scores_gemma":[0.98895,0.005944248,0.0005673437,0.003789423,0.0004465521,0.0003024131],"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":[0.00002264787,0.0001423724,0.000001101488,0.2372348,0.000008278203,0.000005678571,0.00001771269,1.335068e-8,3.25245e-8,0.0001472282,0.1269787,0.6354414],"study_design_scores_gemma":[0.0005566598,0.0001398521,5.31742e-7,0.4384708,0.000005665661,0.000007230094,0.00001728674,0.0005605562,6.331056e-8,0.000478833,0.5595971,0.0001653749],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"protocol","genre_gemma":"protocol","genre_scores_codex":[4.304188e-9,0.2628762,0.0009280067,0.0004343469,0.00001558562,0.7353303,0.0003391001,0.00001966941,0.00005669849],"genre_scores_gemma":[1.317763e-9,0.07948503,0.0106107,0.00003891713,0.0001518379,0.9089934,0.0002968572,0.00003809837,0.0003852163],"genre_candidate":"protocol","genre_consensus":"protocol","teacher_disagreement_score":0.6352761,"threshold_uncertainty_score":0.9999669,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.9494890316243164,"score_gpt":0.7658946204988578,"score_spread":0.1835944111254586,"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."}}