{"id":"W2970992672","doi":"10.14778/3352063.3352116","title":"Data lake management","year":2019,"lang":"en","type":"article","venue":"Proceedings of the VLDB Endowment","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":236,"is_retracted":false,"has_abstract":true,"ca_institutions":"TD Bank Group; University of Toronto","funders":"","keywords":"Metadata; Data management; Metadata management; Data science; Data management plan; Computer science; Data integration; Software versioning; Data mapping; Data extraction; Data element; Research data; Data virtualization; Data curation; Database; World Wide Web; Software","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":["open_science","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.003986794,0.0001340955,0.0002335676,0.0001350364,0.0000818218,0.000221756,0.005522825,0.00002794514,0.0008325573],"category_scores_gemma":[0.0002932887,0.00007762691,0.00008796646,0.0005859726,0.0000711519,0.0006484029,0.005744299,0.00009131345,0.000896293],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002484521,"about_ca_system_score_gemma":0.00000963265,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001333288,"about_ca_topic_score_gemma":0.00002571617,"domain_scores_codex":[0.9966341,0.00001660896,0.0005898306,0.000625041,0.001880161,0.0002542783],"domain_scores_gemma":[0.9979554,0.0001045313,0.0003983375,0.001337419,0.0001476777,0.00005667681],"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":[0.00006041762,0.0002915908,0.01453879,0.0001911041,0.000173583,8.728077e-7,0.0003863852,0.00002188333,0.00112191,0.3760752,0.5550478,0.05209047],"study_design_scores_gemma":[0.0004677484,0.0000371259,0.01089679,0.00005279854,0.00004019351,0.000001624575,0.001976713,0.0002219013,0.001609044,0.03270717,0.9518612,0.0001277166],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.2131897,0.000132185,0.0002619935,0.01205264,0.002108751,0.0025781,0.0005677844,0.00009785569,0.769011],"genre_scores_gemma":[0.9499258,0.00007182606,0.002698324,0.00122611,0.00006681453,0.00003066623,0.00002150233,0.00001461448,0.04594429],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7367362,"threshold_uncertainty_score":0.9998816,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.190965546163535,"score_gpt":0.3896089267691528,"score_spread":0.1986433806056178,"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."}}