{"id":"W4320914459","doi":"10.36227/techrxiv.22060118.v1","title":"DataCurator","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia; Simon Fraser University","funders":"","keywords":"Computer science; Executable; Scalability; Python (programming language); Preprocessor; Verifiable secret sharing; Workflow; Data mining; Scripting language; Programming language; Database; Set (abstract data type)","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":[],"consensus_categories":[],"category_scores_codex":[0.01003315,0.003348063,0.003339498,0.003447136,0.001697595,0.00647915,0.006304326,0.002363418,0.1373182],"category_scores_gemma":[0.03293818,0.003207023,0.003603331,0.002926022,0.001682002,0.006872046,0.008560965,0.006737805,0.1816142],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001802704,"about_ca_system_score_gemma":0.006223147,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002775426,"about_ca_topic_score_gemma":0.00290569,"domain_scores_codex":[0.9930819,0.001200505,0.0008660648,0.001994629,0.002179175,0.0006777277],"domain_scores_gemma":[0.9812904,0.005475434,0.001071585,0.006127425,0.005097066,0.0009381038],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0008452515,0.0001126553,0.002329001,0.001941921,0.0003153595,0.0002764255,0.0007171108,0.0008835262,0.005514437,0.01170705,0.9082642,0.06709308],"study_design_scores_gemma":[0.0002273004,0.00006551699,0.001647546,0.0003029538,0.00008554762,0.0002584037,0.0001193961,0.005638674,0.01281868,0.0190802,0.959595,0.0001607117],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"software","genre_gemma":"software","genre_scores_codex":[0.001112511,0.0004414185,0.1689291,0.0006926672,0.0005415526,0.0006311081,0.07013289,0.7447227,0.01279621],"genre_scores_gemma":[0.01627703,0.0008311478,0.1832402,0.00310092,0.0002965052,0.003996219,0.2578175,0.5026225,0.031818],"genre_candidate":"software","genre_consensus":"software","teacher_disagreement_score":0.1373182,"threshold_uncertainty_score":0.4593752,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.6702760272205386,"score_gpt":0.5217964719834293,"score_spread":0.1484795552371093,"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."}}