{"id":"W4399174293","doi":"10.1145/3654921","title":"Reservoir Sampling over Joins","year":2024,"lang":"en","type":"article","venue":"Proceedings of the ACM on Management of Data","topic":"Data Stream Mining Techniques","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Joins; Tuple; Computer science; Join (topology); Graph; Theoretical computer science; Sampling (signal processing); Operator (biology); Data mining; Mathematics; Discrete mathematics","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":["open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.0007913063,0.0001225803,0.0001387795,0.0001549722,0.00004645425,0.00021136,0.01753785,0.000028986,0.000007797756],"category_scores_gemma":[0.0002213357,0.00008835139,0.00004654874,0.0006329475,0.00005640659,0.001413268,0.02627618,0.0001275121,0.000009249996],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002409232,"about_ca_system_score_gemma":0.00001135436,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001236989,"about_ca_topic_score_gemma":4.109657e-7,"domain_scores_codex":[0.9984995,0.000004093562,0.0002636488,0.0005338515,0.0005225379,0.0001763938],"domain_scores_gemma":[0.9964963,0.00006712403,0.0001372459,0.003226153,0.0000478999,0.00002526079],"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":[0.00001225787,0.00010128,0.0006106788,0.001328099,0.0001790521,0.000002752041,0.0001522855,0.00000223844,0.002167499,0.5522776,0.3901671,0.05299922],"study_design_scores_gemma":[0.0006327116,0.0005233621,0.01679902,0.01016505,0.0003162784,0.00001255231,0.0003674848,0.02971886,0.06761117,0.4356281,0.4373125,0.0009129881],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.5219773,0.00223277,0.1143882,0.08472753,0.003773032,0.005495519,0.002703296,0.006583806,0.2581185],"genre_scores_gemma":[0.25835,0.0002050345,0.7405537,0.0001528861,0.00005383756,0.00001547874,0.00002589906,0.00002379924,0.0006193818],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6261654,"threshold_uncertainty_score":0.9877777,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.12639640691388,"score_gpt":0.35644566552144,"score_spread":0.23004925860756,"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."}}