{"id":"W2103615139","doi":"10.14778/1453856.1453955","title":"Keyword query cleaning","year":2008,"lang":"en","type":"article","venue":"Proceedings of the VLDB Endowment","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":51,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University; Ontario Tech University","funders":"","keywords":"Computer science; Query optimization; Query expansion; Web query classification; Web search query; Sargable; Query language; Information retrieval; Online aggregation; View; Set (abstract data type); Context (archaeology); Database; Query by Example; Data mining; Search engine","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":[0.004937681,0.001673253,0.003820275,0.003554981,0.002591608,0.004711909,0.004240023,0.002464984,0.005817516],"category_scores_gemma":[0.02934631,0.001033707,0.002603736,0.00661949,0.001608407,0.008339137,0.005772573,0.002531371,0.005920553],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001356313,"about_ca_system_score_gemma":0.003843148,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003232518,"about_ca_topic_score_gemma":0.002473797,"domain_scores_codex":[0.9873661,0.00238954,0.001919977,0.002775749,0.004555046,0.0009936746],"domain_scores_gemma":[0.9793956,0.005963549,0.001222691,0.007473981,0.005598207,0.0003460422],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001572318,0.0004880342,0.007349462,0.002452458,0.000446247,0.001091942,0.00178557,0.02608369,0.07403389,0.03928134,0.06947888,0.7759361],"study_design_scores_gemma":[0.000314926,0.0008275412,0.005487939,0.0004726626,0.0006274961,0.008258673,0.004474565,0.3924557,0.2026276,0.1363371,0.2476384,0.0004774226],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01959717,0.002480983,0.9621047,0.0007570855,0.0002858732,0.0006551513,0.002136894,0.007421462,0.004560646],"genre_scores_gemma":[0.18961,0.001962342,0.7898313,0.001141436,0.0002727875,0.000602523,0.007815471,0.002445092,0.006319128],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005817516,"threshold_uncertainty_score":0.02611327,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01646142064894577,"score_gpt":0.1908917334095587,"score_spread":0.1744303127606129,"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."}}