{"id":"W3146147323","doi":"","title":"Lakehouse: A New Generation of Open Platforms that Unify Data Warehousing and Advanced Analytics.","year":2021,"lang":"en","type":"article","venue":"Conference on Innovative Data Systems Research","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":123,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Analytics; Data warehouse; Data science; Database","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":["metaresearch","scholarly_communication","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.01920787,0.0001846625,0.0005634697,0.0005990985,0.0003330643,0.003047313,0.007935701,0.00009173081,0.000237139],"category_scores_gemma":[0.008996663,0.0001400913,0.00001142573,0.004048625,0.0003438755,0.004107798,0.01728054,0.0004329756,0.00007979927],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000617585,"about_ca_system_score_gemma":0.001479583,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00130949,"about_ca_topic_score_gemma":0.001838082,"domain_scores_codex":[0.9932848,0.0007882873,0.001013928,0.001601335,0.002890029,0.0004215709],"domain_scores_gemma":[0.9875896,0.001291829,0.0004681597,0.007103241,0.003382313,0.000164864],"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":[0.0001245676,0.0001745498,0.001017623,0.0001318543,0.0001444601,0.00005834777,0.0007685958,0.00007274096,0.003483747,0.4615848,0.3228096,0.2096291],"study_design_scores_gemma":[0.003688337,0.0005489678,0.003738967,0.001484644,0.00003481059,0.00003717185,0.06252182,0.1739665,0.006345469,0.0286581,0.7179775,0.0009976816],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.249746,0.002770397,0.5759263,0.02307307,0.003440607,0.009840389,0.06722741,0.0001945618,0.06778129],"genre_scores_gemma":[0.9801367,0.0003758504,0.002712554,0.0001317854,0.0001367286,0.00001900054,0.01026055,0.00002079907,0.006206062],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7303907,"threshold_uncertainty_score":0.999351,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.946872203848267,"score_gpt":0.6271471361658597,"score_spread":0.3197250676824073,"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."}}