{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005297637,0.001098948,0.00114686,0.001970623,0.001367452,0.003220745,0.003376425,0.000937036,0.01977639],"category_scores_gemma":[0.01372293,0.0007492726,0.001746283,0.002257868,0.001221177,0.004726538,0.006444715,0.001458352,0.005891527],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006244694,"about_ca_system_score_gemma":0.001737375,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001621153,"about_ca_topic_score_gemma":0.002143116,"domain_scores_codex":[0.9921705,0.001029798,0.000722334,0.001950142,0.003713248,0.0004139831],"domain_scores_gemma":[0.9898599,0.002834718,0.0004376196,0.004863498,0.001716795,0.0002874784],"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":[0.001457819,0.0004875447,0.01458256,0.0008799261,0.0003329797,0.0005058463,0.001251333,0.0238037,0.02243548,0.09813234,0.0789342,0.7571962],"study_design_scores_gemma":[0.0002826265,0.0004530248,0.003982882,0.0001628235,0.0001795885,0.001742434,0.0008230765,0.4322891,0.0912307,0.1832447,0.2854356,0.000173437],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01376038,0.0003608255,0.9247355,0.000210331,0.0001704814,0.0003939779,0.002495207,0.04491507,0.01295812],"genre_scores_gemma":[0.2091268,0.0003621225,0.7531769,0.0005445711,0.0001805557,0.000528625,0.01361135,0.008758059,0.01371111],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01977639,"threshold_uncertainty_score":0,"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."}}