{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.03199944,0.0005148244,0.001100996,0.009278895,0.001026327,0.003781405,0.001671698,0.0008429628,0.002273497],"category_scores_gemma":[0.1269387,0.0004363851,0.001149837,0.0101251,0.0009687205,0.004024926,0.002243966,0.0007387569,0.0005537609],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001786037,"about_ca_system_score_gemma":0.001321378,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002958758,"about_ca_topic_score_gemma":0.003838259,"domain_scores_codex":[0.9829919,0.005877574,0.00347141,0.002134588,0.00517271,0.0003517939],"domain_scores_gemma":[0.7990964,0.1363451,0.01199754,0.03141007,0.0202441,0.0009066911],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.003934097,0.0006351763,0.3384908,0.008443042,0.002481352,0.001005059,0.006982224,0.06006933,0.06034914,0.03507587,0.04813419,0.4343997],"study_design_scores_gemma":[0.0005003409,0.001361648,0.3739217,0.0007563782,0.001432213,0.003167904,0.005617525,0.3045124,0.135409,0.04758358,0.1251774,0.0005599602],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6595489,0.00253599,0.2149938,0.002169919,0.0001804783,0.0004491454,0.09476496,0.01844915,0.006907814],"genre_scores_gemma":[0.6874428,0.000801631,0.1815171,0.0003292153,0.00007771218,0.0005260374,0.1268108,0.001306369,0.001188358],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9680006,"threshold_uncertainty_score":0.1692312,"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."}}