{"id":"W2162672362","doi":"10.1109/wiiat.2008.24","title":"An Unsupervised Approach to Cluster Web Search Results Based on Word Sense Communities","year":2008,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Cluster analysis; Word (group theory); Information retrieval; Measure (data warehouse); Process (computing); Modularity (biology); Web page; Artificial intelligence; Unsupervised learning; Data mining; World Wide Web","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":[0.002149212,0.0008853294,0.0009217185,0.01053931,0.001863624,0.001614845,0.001628229,0.001066349,0.0007027157],"category_scores_gemma":[0.009886341,0.0005881282,0.001333665,0.006438124,0.00151725,0.003062252,0.001807834,0.001087848,0.0005355112],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001039569,"about_ca_system_score_gemma":0.001999424,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005033613,"about_ca_topic_score_gemma":0.01388543,"domain_scores_codex":[0.9974161,0.0007199423,0.0001968768,0.0007305701,0.000799006,0.0001374043],"domain_scores_gemma":[0.9942513,0.001906342,0.001071254,0.0007870489,0.001747927,0.0002363291],"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":[0.0004043869,0.000989418,0.02725777,0.0008872965,0.000789369,0.0004959653,0.004239549,0.05157814,0.04371579,0.06073567,0.01083122,0.7980756],"study_design_scores_gemma":[0.000103176,0.0002694873,0.01424367,0.0001607038,0.0002600081,0.001250541,0.001292929,0.8467529,0.01783461,0.1064096,0.01117737,0.0002448577],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0262277,0.0004503436,0.9703658,0.0002447606,0.0000432085,0.0003470893,0.000232639,0.0007708013,0.001317644],"genre_scores_gemma":[0.2306935,0.000288659,0.7657697,0.0001244301,0.0001127401,0.000436293,0.0006878452,0.0001432867,0.001743619],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01053931,"threshold_uncertainty_score":0.01136625,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06940130893529436,"score_gpt":0.2719062895478437,"score_spread":0.2025049806125494,"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."}}