{"id":"W4283583762","doi":"10.20944/preprints202206.0335.v1","title":"The Dataharmonizer: a Tool for Faster Data Harmonization, Validation, Aggregation, and Analysis of Pathogen Genomics Contextual Information","year":2022,"lang":"en","type":"preprint","venue":"Preprints.org","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"St. John’s Health Sciences Centre; BC Centre for Disease Control; McMaster University; Ottawa Public Health; Nova Scotia Health Authority; Simon Fraser University; Public Health Agency of Canada; Hospital for Sick Children; University of British Columbia; Public Health Ontario; University of Alberta; Saskatchewan Disease Control Laboratory; Institut National de Santé Publique du Québec","funders":"","keywords":"Metadata; Data sharing; Data science; Harmonization; Computer science; Interoperability; Big data; Data integration; Contextual design; World Wide Web; Database; Data mining; Medicine","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.02257993,0.004107668,0.002752478,0.008724011,0.001921162,0.008072344,0.00534957,0.002130759,0.03189761],"category_scores_gemma":[0.04133322,0.003083575,0.003288149,0.006400321,0.001973029,0.01050123,0.01267186,0.006079113,0.01647471],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001970007,"about_ca_system_score_gemma":0.005366955,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006066946,"about_ca_topic_score_gemma":0.005854341,"domain_scores_codex":[0.9879249,0.002355102,0.001942929,0.002403961,0.004814545,0.0005584336],"domain_scores_gemma":[0.9729385,0.01400743,0.001552175,0.007457999,0.002813684,0.001230142],"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":[0.001896641,0.000519715,0.00902897,0.003072529,0.0008254991,0.001537365,0.003892733,0.009364299,0.01507402,0.03439352,0.4681377,0.452257],"study_design_scores_gemma":[0.0009800191,0.0002955218,0.008396921,0.001823085,0.0002472867,0.001004024,0.001079027,0.1008278,0.04725319,0.07316355,0.7641308,0.0007987339],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.002761321,0.0003788781,0.490229,0.0008728976,0.0003183272,0.0008311013,0.03023229,0.4684369,0.005939198],"genre_scores_gemma":[0.03542018,0.0007456536,0.784922,0.00171977,0.0001293579,0.002223067,0.08236159,0.0843808,0.008097577],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.03189761,"threshold_uncertainty_score":0.1194155,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.321146004430793,"score_gpt":0.4154555473716058,"score_spread":0.09430954294081278,"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."}}