{"id":"W2746538584","doi":"10.3897/tdwgproceedings.1.20380","title":"Using YesWorkflow hybrid queries to reveal data lineage from data curation activities","year":2017,"lang":"en","type":"article","venue":"Biodiversity Information Science and Standards","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Agriculture and Agri-Food Canada","funders":"National Science Foundation","keywords":"Workflow; Computer science; Scripting language; Metadata; Python (programming language); Database; Annotation; World Wide Web; Information retrieval; Programming language","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":[0.01024879,0.001074204,0.0009858932,0.004412541,0.001896593,0.006351479,0.001911914,0.001713658,0.007312152],"category_scores_gemma":[0.03221295,0.000919856,0.0009565947,0.003333794,0.001161243,0.009868728,0.005840777,0.001628818,0.001874056],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001268316,"about_ca_system_score_gemma":0.002690898,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004424358,"about_ca_topic_score_gemma":0.004383781,"domain_scores_codex":[0.9934622,0.001149668,0.001063469,0.001341885,0.002629565,0.0003531523],"domain_scores_gemma":[0.9699069,0.01507614,0.001668746,0.008959172,0.00339786,0.0009911349],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.006355041,0.0009936832,0.1210349,0.001440566,0.0006247747,0.002047038,0.007436384,0.02175045,0.06298884,0.08004125,0.1847567,0.5105304],"study_design_scores_gemma":[0.0003982088,0.000421986,0.01741924,0.0003712663,0.0002765262,0.0008160621,0.002276326,0.5256119,0.1157223,0.1042319,0.2321021,0.0003521581],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1445131,0.0005667446,0.6285229,0.003509827,0.0009091825,0.000865956,0.02754056,0.1773255,0.01624618],"genre_scores_gemma":[0.4873667,0.0002934338,0.4339501,0.001304131,0.0003000339,0.0007302727,0.04934895,0.01662589,0.01008048],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01024879,"threshold_uncertainty_score":0.05420136,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3728313323745507,"score_gpt":0.4466233665384567,"score_spread":0.073792034163906,"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."}}