{"id":"W4400875444","doi":"10.3390/computers13070183","title":"Data Lakes: A Survey of Concepts and Architectures","year":2024,"lang":"en","type":"article","venue":"Computers","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"Ministry of Higher Education and Scientific Research","keywords":"Data science; Survey data collection; Computer science; Geography; Mathematics; Statistics","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":[0.007764793,0.001251201,0.001110993,0.01617883,0.00156167,0.007792491,0.002715487,0.001808506,0.004548352],"category_scores_gemma":[0.01282962,0.001406062,0.001370153,0.02854063,0.003751261,0.024213,0.004379464,0.003026724,0.001397733],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00415775,"about_ca_system_score_gemma":0.005514023,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005504338,"about_ca_topic_score_gemma":0.004119903,"domain_scores_codex":[0.9939016,0.001690231,0.001220186,0.0007817816,0.002103926,0.0003023446],"domain_scores_gemma":[0.9888853,0.007456791,0.0008521093,0.0007449786,0.001772121,0.0002887079],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00008607782,0.00006032328,0.002828982,0.01001922,0.0001076058,0.0002739373,0.003534013,0.00402706,0.001619568,0.4820082,0.02979806,0.465637],"study_design_scores_gemma":[0.00001029441,0.00005126576,0.001372855,0.005628515,0.0000602614,0.0008600071,0.00208755,0.002908709,0.001025228,0.1015171,0.8844087,0.00006949864],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.01246486,0.6303055,0.2928607,0.0149971,0.001010118,0.0003220758,0.002426732,0.00178057,0.04383236],"genre_scores_gemma":[0.06411577,0.6800381,0.2409626,0.003198333,0.0008768348,0.0005293222,0.003633102,0.0006120085,0.006033948],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.01617883,"threshold_uncertainty_score":0.04106462,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3904131326033246,"score_gpt":0.4888631919173062,"score_spread":0.09845005931398154,"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."}}