{"id":"W2022006917","doi":"10.1145/1645953.1646069","title":"Clustering web queries","year":2009,"lang":"en","type":"article","venue":"","topic":"Advanced Clustering Algorithms Research","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Cluster analysis; Computer science; Data mining; Weighting; Classifier (UML); Set (abstract data type); Population; Search engine; Information retrieval; Metric (unit); Artificial intelligence","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":[0.005239871,0.001459625,0.001473242,0.01039758,0.001502833,0.003605348,0.001778453,0.001484199,0.001175991],"category_scores_gemma":[0.02291179,0.0003347423,0.001321454,0.006619292,0.000658394,0.002486611,0.001068729,0.0008734359,0.001054675],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00228193,"about_ca_system_score_gemma":0.001537607,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01047698,"about_ca_topic_score_gemma":0.01230538,"domain_scores_codex":[0.9896288,0.002486056,0.001045799,0.002278811,0.004092177,0.0004684131],"domain_scores_gemma":[0.9841325,0.005660839,0.001015368,0.003125167,0.005789793,0.0002763642],"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":[0.001046757,0.00120516,0.1030385,0.00145043,0.0009951032,0.0003160553,0.002040006,0.09833182,0.03522867,0.011507,0.02609568,0.7187448],"study_design_scores_gemma":[0.000104006,0.0005999582,0.07392479,0.0001788875,0.0003052336,0.001320006,0.002310463,0.8230383,0.04919229,0.02138843,0.02740704,0.0002305671],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4517521,0.003122724,0.5162309,0.0008686598,0.0002694704,0.002615649,0.008343474,0.006023019,0.01077401],"genre_scores_gemma":[0.5979475,0.0006418353,0.3789138,0.0002359231,0.0001674048,0.0006054989,0.01717752,0.0005495529,0.003760896],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01047698,"threshold_uncertainty_score":0.02771145,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01887784931709801,"score_gpt":0.3006797422883326,"score_spread":0.2818018929712346,"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."}}