{"id":"W6950346504","doi":"10.5281/zenodo.7998268","title":"Successes, pain points, and lessons learned when curating data at scale: The NACJD experience","year":2023,"lang":"en","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Data Analysis and Archiving","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Institute for Public Safety Research and Treatment","funders":"","keywords":"Workflow; Restructuring; Criminal justice; Economic Justice; Presentation (obstetrics); The Internet; Period (music); Data curation","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":"codex-gemma-dda1882f352a","candidate_categories":["sts","scholarly_communication","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.004347514,0.00008579081,0.0001018561,0.0001008791,0.008433498,0.001437699,0.001971574,0.00002893872,0.004254733],"category_scores_gemma":[0.003724576,0.00007182371,0.0000224029,0.0007587383,0.0004888527,0.000593219,0.004112307,0.0001657328,0.001536308],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005150789,"about_ca_system_score_gemma":0.000006430752,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005003986,"about_ca_topic_score_gemma":0.0001335969,"domain_scores_codex":[0.9972883,0.001254053,0.000175934,0.0004450514,0.0004621735,0.0003744749],"domain_scores_gemma":[0.998801,0.0001963579,0.00008352644,0.0006515174,0.0001253617,0.0001422176],"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":[0.00001529523,0.00003498583,0.0002966296,0.00003189243,0.00003697853,0.00001261214,0.05388745,0.00001037656,0.00161069,0.01728023,0.4071365,0.5196463],"study_design_scores_gemma":[0.0000807123,0.00001545378,0.001011162,0.00002204989,0.000008281972,0.000003687803,0.01257102,0.002005177,0.00001929181,0.0004604287,0.9837036,0.00009910586],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.4400595,0.000556217,0.008881695,0.1931716,0.0003639412,0.001609091,0.00244193,0.004302243,0.3486138],"genre_scores_gemma":[0.9892601,0.0009899061,0.0003127592,0.0003814208,0.0002612533,9.650774e-8,0.003178631,0.0004868959,0.005128967],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5765671,"threshold_uncertainty_score":0.9995989,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1384627393882129,"score_gpt":0.3563165711805265,"score_spread":0.2178538317923136,"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."}}