{"id":"W2044921851","doi":"10.1145/2661829.2662010","title":"Computing Multi-Relational Sufficient Statistics for Large Databases","year":2014,"lang":"en","type":"preprint","venue":"","topic":"Bayesian Modeling and Causal Inference","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"Natural Sciences and Engineering Research Council of Canada; China Scholarship Council","keywords":"Computer science; Relational database; Database; Information retrieval; Data mining; Data science","routes":{"ca_aff":true,"ca_fund":true,"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":[0.01273957,0.001193509,0.002439496,0.006716802,0.001485632,0.005317678,0.002902691,0.001334222,0.002099219],"category_scores_gemma":[0.1027583,0.001809481,0.001780712,0.005712016,0.002226027,0.01210961,0.003680782,0.0025954,0.0009103505],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001800036,"about_ca_system_score_gemma":0.004093711,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002471642,"about_ca_topic_score_gemma":0.004440141,"domain_scores_codex":[0.9871012,0.004110276,0.001802136,0.002269507,0.004238745,0.0004781127],"domain_scores_gemma":[0.9055502,0.07283483,0.005553974,0.009033989,0.005276245,0.001750634],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0007713982,0.000447671,0.03371787,0.001217512,0.0006266508,0.001103192,0.001096449,0.3828462,0.00796234,0.3297202,0.0103928,0.2300977],"study_design_scores_gemma":[0.00003420175,0.0000448467,0.0009499487,0.00004607219,0.00004204826,0.0001355522,0.0001245741,0.7573504,0.002569965,0.2368701,0.0018049,0.00002742005],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0341225,0.0004553311,0.9601062,0.0006237879,0.00003319671,0.00007887546,0.001491226,0.002463866,0.0006250625],"genre_scores_gemma":[0.4333244,0.0005344693,0.5582218,0.0002734262,0.0001802207,0.0004138928,0.006065298,0.000411557,0.0005749968],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01273957,"threshold_uncertainty_score":0.06737411,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1120781166882575,"score_gpt":0.3458678401406248,"score_spread":0.2337897234523673,"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."}}