{"id":"W2399480695","doi":"","title":"Summary-based Comparison of Data Quality across Public MAGE-ML Genomic Datasets","year":2010,"lang":"en","type":"article","venue":"Cadernos de Linguística e Teoria da Literatura (Universidade Federal de Minas Gerais)","topic":"Genomics and Phylogenetic Studies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Schema (genetic algorithms); Data mining; XML; Reuse; Microarray databases; Microarray analysis techniques; Schema evolution; Data science; Information retrieval; World Wide Web; Database schema; Biology","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0007362781,0.0004231543,0.000531629,0.0001004369,0.0004653278,0.000527549,0.001471931,0.0004875355,0.00005197355],"category_scores_gemma":[0.0005508316,0.0004634483,0.0001703477,0.0002070236,0.0004280654,0.00001721535,0.0009847095,0.0006271983,0.0000107725],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005421187,"about_ca_system_score_gemma":0.0004148268,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000131713,"about_ca_topic_score_gemma":0.0001630847,"domain_scores_codex":[0.9971747,0.0002063943,0.0005559431,0.0009515038,0.0002507767,0.0008606689],"domain_scores_gemma":[0.9972389,0.0001785267,0.0003383447,0.001652132,0.0002425777,0.0003494609],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002750061,0.0001980029,0.01892978,0.000119914,0.0001856508,0.00002571526,0.0009421877,0.0000508713,0.9693944,0.001624111,0.006005569,0.002248803],"study_design_scores_gemma":[0.00636693,0.001011169,0.1061534,0.0001365531,0.0003298534,0.0001212961,0.003618881,0.01111975,0.09265648,0.0008752979,0.7749691,0.002641271],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9893817,0.001446886,0.00182622,0.0007806294,0.0005410123,0.0002981799,0.00521551,0.00002187743,0.0004879353],"genre_scores_gemma":[0.9775286,0.00002968258,0.01635304,0.0003863111,0.000646211,0.000007963506,0.00473262,0.0000515619,0.0002639728],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8767379,"threshold_uncertainty_score":0.9997817,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04737643444670213,"score_gpt":0.3459348718103532,"score_spread":0.298558437363651,"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."}}