{"id":"W3087826863","doi":"10.48550/arxiv.2009.12415","title":"A Big Data Lake for Multilevel Streaming Analytics","year":2020,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"Ontario Centres of Excellence","keywords":"Computer science; Big data; Analytics; Data warehouse; Data science; Data management; Scalability; Raw data; Volume (thermodynamics); Data analysis; Database; Unstructured data; Variety (cybernetics); Architecture; Business intelligence; Data mining","routes":{"ca_aff":true,"ca_fund":true,"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":[0.007047119,0.0007721713,0.0006017255,0.002150869,0.002407447,0.006277382,0.002961572,0.001271133,0.006911325],"category_scores_gemma":[0.01637147,0.0009715799,0.0009479261,0.003122955,0.001604147,0.01069829,0.008109072,0.003294168,0.002763907],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001791288,"about_ca_system_score_gemma":0.004468654,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003609587,"about_ca_topic_score_gemma":0.005430643,"domain_scores_codex":[0.9958065,0.0007272247,0.0003763711,0.0005567948,0.002263852,0.0002691604],"domain_scores_gemma":[0.9900074,0.001833075,0.0004550202,0.002828646,0.003442595,0.001433166],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001307819,0.000599349,0.0159249,0.0008987634,0.0003174939,0.001436854,0.002934761,0.02694352,0.03371562,0.1446639,0.4043422,0.3669148],"study_design_scores_gemma":[0.0003090647,0.000481068,0.005249904,0.0003844106,0.00008846166,0.0005298438,0.001820827,0.3411602,0.02721632,0.1731534,0.4493995,0.0002070937],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02833606,0.001027416,0.8508818,0.01673217,0.001109758,0.001526703,0.00624168,0.07152594,0.0226184],"genre_scores_gemma":[0.152443,0.0006582292,0.8240207,0.00130818,0.0003728747,0.0006872594,0.0102887,0.002735933,0.007485048],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007047119,"threshold_uncertainty_score":0.03726918,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.7545566830413366,"score_gpt":0.3475097281097882,"score_spread":0.4070469549315484,"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."}}