{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01351263,0.001064766,0.001422133,0.007892552,0.003122945,0.01269661,0.005567789,0.00149881,0.0280816],"category_scores_gemma":[0.0414289,0.001078477,0.001497977,0.009107835,0.001126833,0.01718519,0.01127162,0.00326425,0.01830473],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002542973,"about_ca_system_score_gemma":0.007144351,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006534911,"about_ca_topic_score_gemma":0.004540147,"domain_scores_codex":[0.9902188,0.001388216,0.001680583,0.001443264,0.004585288,0.0006839008],"domain_scores_gemma":[0.9779451,0.003213963,0.001534096,0.009015027,0.006640497,0.001651296],"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":[0.0002141557,0.00008546647,0.006912315,0.0009170364,0.0001310364,0.0002502174,0.00113894,0.002427701,0.003093446,0.07858947,0.5607018,0.3455383],"study_design_scores_gemma":[0.00003398857,0.00003280322,0.001850776,0.0002551512,0.00004663851,0.0002070739,0.0004127493,0.01122277,0.005973069,0.0329357,0.9469594,0.00006990021],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01169187,0.007111954,0.5594096,0.02324033,0.002832851,0.003080809,0.1010732,0.1775492,0.1140102],"genre_scores_gemma":[0.1118763,0.007822782,0.5482424,0.005196006,0.001630319,0.002458069,0.222079,0.02403266,0.07666242],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0280816,"threshold_uncertainty_score":0.09394228,"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."}}