{"id":"W2616147950","doi":"10.14778/3115404.3115409","title":"Auto-join","year":2017,"lang":"en","type":"article","venue":"Proceedings of the VLDB Endowment","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":66,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Joins; Join (topology); Computer science; Transformation (genetics); Domain (mathematical analysis); String (physics); Database; Data mining; Theoretical computer science; 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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00326452,0.0001059617,0.0002036543,0.00007488864,0.000509723,0.0006620825,0.003779837,0.00003061836,0.0001923004],"category_scores_gemma":[0.002977339,0.00005891431,0.0001404079,0.0001214365,0.0002163719,0.0006064652,0.002001446,0.00008723116,0.0001954577],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003134917,"about_ca_system_score_gemma":0.00001673951,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001157699,"about_ca_topic_score_gemma":0.00001310102,"domain_scores_codex":[0.9976644,0.00001054078,0.0004533225,0.0003297332,0.001331779,0.0002102305],"domain_scores_gemma":[0.9981151,0.00008276365,0.0007563882,0.000758498,0.000225204,0.00006204531],"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":[0.00004967026,0.0002360422,0.01742536,0.0000561211,0.00007307003,8.772079e-7,0.001178971,0.000004912488,0.01239709,0.491233,0.384108,0.09323694],"study_design_scores_gemma":[0.000665396,0.00007226803,0.08386739,0.0000710882,0.00003837509,0.00000325142,0.001533925,0.0002289126,0.05437285,0.2421853,0.6167632,0.0001979737],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.3593015,0.0000888122,0.0002548805,0.04407267,0.001923103,0.001033821,0.00005741832,0.00007551115,0.5931923],"genre_scores_gemma":[0.9845093,0.00001609309,0.0007370529,0.0004202429,0.00006947974,0.00002097857,3.231767e-7,0.000005812296,0.01422071],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6252078,"threshold_uncertainty_score":0.702394,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1996264917742379,"score_gpt":0.4134582450415684,"score_spread":0.2138317532673306,"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."}}