{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001167686,0.00008204108,0.00008353345,0.00007566743,0.0000890374,0.0001487541,0.0007075295,0.00002667194,0.00002676326],"category_scores_gemma":[0.0000297955,0.0000721082,0.00002717478,0.0002767969,0.00002205024,0.0005763689,0.0002692395,0.0001000848,0.0001157431],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003749249,"about_ca_system_score_gemma":0.00003617302,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004603606,"about_ca_topic_score_gemma":0.000006218377,"domain_scores_codex":[0.9990824,0.00001924235,0.0001086184,0.0002493906,0.0002362046,0.0003040961],"domain_scores_gemma":[0.9993656,0.00003097947,0.00001700339,0.0004554377,0.00004386211,0.00008707704],"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":[0.000005882548,0.0000412407,0.000035271,0.000007568364,0.000005200134,0.00006499622,0.0003172494,0.00201512,0.01033007,0.03673571,0.0009473309,0.9494944],"study_design_scores_gemma":[0.000253553,0.0001653884,0.001186737,0.00001172904,4.875187e-7,0.00005741749,0.00002647153,0.9737675,0.004848312,0.006192479,0.01327202,0.0002178834],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.001222767,0.00003795468,0.9655281,0.002678547,0.0001127,0.00006010736,1.976713e-7,0.0004687513,0.02989085],"genre_scores_gemma":[0.4581539,0.00001548795,0.5367584,0.0006357689,0.00005772807,0.000004038936,2.523302e-7,0.000004713268,0.004369782],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9717524,"threshold_uncertainty_score":0.2940487,"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."}}