{"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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.002131118,0.0003126845,0.0005548364,0.0003764998,0.000197308,0.0004163975,0.006241909,0.0002166363,0.0001271177],"category_scores_gemma":[0.002774646,0.0003155708,0.0002429529,0.000696949,0.0001441653,0.0004297606,0.01127314,0.0003858498,0.0002872203],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000685432,"about_ca_system_score_gemma":0.0002160937,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000146632,"about_ca_topic_score_gemma":0.002082733,"domain_scores_codex":[0.9963257,0.0002023548,0.0005249435,0.002223296,0.0003677924,0.000355939],"domain_scores_gemma":[0.9936922,0.001180291,0.0005492679,0.004058061,0.0002774337,0.0002427427],"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":[0.0006156524,0.0007740458,0.006656996,0.0007241739,0.001516158,0.0008645057,0.001163319,0.1695743,0.0000241784,0.3007745,0.3242247,0.1930875],"study_design_scores_gemma":[0.0006006773,0.00003476598,0.00124906,0.00004736143,0.0002755703,3.893185e-7,0.001038695,0.6098276,0.000006378878,0.1163379,0.2701719,0.0004097157],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01216112,0.00002035285,0.9735736,0.0007990405,0.001208825,0.0006785762,0.008738223,0.0001170533,0.002703191],"genre_scores_gemma":[0.9866648,0.00008754679,0.002695878,0.0004795989,0.0003260155,0.000001258467,0.001790985,0.00002387952,0.007930079],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9745036,"threshold_uncertainty_score":0.9999297,"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."}}