{"id":"W2260484439","doi":"10.14778/2824032.2824036","title":"SEMA-JOIN","year":2015,"lang":"en","type":"article","venue":"Proceedings of the VLDB Endowment","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":31,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Joins; Join (topology); Hash join; Table (database); Theoretical computer science; Data mining; Database; Programming language; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006727853,0.001161083,0.001338868,0.002542997,0.001764882,0.00444622,0.00381952,0.00121987,0.01859626],"category_scores_gemma":[0.01381962,0.0009098294,0.00242496,0.003420101,0.001317569,0.005640713,0.005851528,0.002124039,0.008025618],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007420597,"about_ca_system_score_gemma":0.002101909,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001601564,"about_ca_topic_score_gemma":0.003306408,"domain_scores_codex":[0.9920025,0.001477284,0.0008342109,0.001798519,0.003568155,0.0003193906],"domain_scores_gemma":[0.9918619,0.002995902,0.000578505,0.003138031,0.001070841,0.000354697],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001367992,0.0005552369,0.01036199,0.001324896,0.0005683992,0.0005322447,0.001211659,0.02552644,0.01620577,0.2049136,0.1906248,0.546807],"study_design_scores_gemma":[0.0002060314,0.0003583563,0.002543492,0.0001563545,0.000113222,0.001222432,0.0005133785,0.3132018,0.02759614,0.2355093,0.4184374,0.0001420098],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005088016,0.0003526405,0.9388716,0.0003708552,0.0002352964,0.0004191787,0.005992004,0.03971449,0.008955968],"genre_scores_gemma":[0.06233814,0.0002828435,0.9093468,0.0005081084,0.000202194,0.000596797,0.01666791,0.003846786,0.006210474],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01859626,"threshold_uncertainty_score":0,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2932342444414948,"score_gpt":0.4016846612147633,"score_spread":0.1084504167732685,"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."}}