{"id":"W2163814946","doi":"10.1145/1557670.1557681","title":"Keyword query cleaning using hidden Markov models","year":2009,"lang":"en","type":"article","venue":"","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ontario Institute of Technology","funders":"","keywords":"Computer science; Hidden Markov model; Query optimization; Probabilistic logic; Web query classification; Data mining; Information retrieval; Keyword spotting; Web search query; Artificial intelligence; 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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001881426,0.00009665683,0.00009259927,0.00009217749,0.00009394722,0.0003486258,0.0008031671,0.00002346031,0.00002719273],"category_scores_gemma":[0.00000406428,0.00008682906,0.00003743147,0.0002702174,0.000009240794,0.002105786,0.000288043,0.00005278435,0.0000324048],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002129416,"about_ca_system_score_gemma":0.00001379867,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000324268,"about_ca_topic_score_gemma":0.000002708655,"domain_scores_codex":[0.9991114,0.00001839663,0.0001352316,0.0002893298,0.000196822,0.0002487863],"domain_scores_gemma":[0.9994163,0.00001297782,0.0000364468,0.0004583363,0.0000203682,0.0000555599],"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":[0.000001529399,0.00002972496,0.00002573888,0.00000272944,0.000007399989,0.0000260308,0.00009395184,0.0003838702,0.0001246453,0.2173691,0.006054235,0.7758811],"study_design_scores_gemma":[0.0001143594,0.00002286886,0.0001781608,0.00001107413,0.000003546746,0.00000393687,0.00002167139,0.9698299,0.0001099709,0.028065,0.001488267,0.0001513133],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.002945991,0.00002918332,0.9104059,0.0006754341,0.0001656381,0.00006522229,4.722949e-7,0.0002534903,0.08545866],"genre_scores_gemma":[0.2324078,0.00001298286,0.7618681,0.001755944,0.0001150404,7.671826e-7,0.000003354538,0.000005703492,0.00383025],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9694459,"threshold_uncertainty_score":0.3540786,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0369287846093509,"score_gpt":0.2555249096461961,"score_spread":0.2185961250368452,"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."}}