{"id":"W1965231990","doi":"10.1145/2030376.2030391","title":"Clustering for semi-supervised spam filtering","year":2011,"lang":"en","type":"article","venue":"","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Cluster analysis; Medoid; Filter (signal processing); Data mining; Artificial intelligence; Set (abstract data type); Training set; Bag-of-words model; Partition (number theory); Pattern recognition (psychology); Machine learning; Mathematics","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":[0.005319247,0.001078421,0.001914387,0.002952729,0.00142367,0.001482885,0.002105143,0.002255664,0.001667062],"category_scores_gemma":[0.01939897,0.0006764106,0.001501295,0.001688066,0.002266448,0.00232895,0.001972656,0.001816122,0.001312413],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001522767,"about_ca_system_score_gemma":0.001187581,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002159839,"about_ca_topic_score_gemma":0.002331306,"domain_scores_codex":[0.994225,0.002990667,0.0003024707,0.001094481,0.001197633,0.000189744],"domain_scores_gemma":[0.9831177,0.008535887,0.001652354,0.003578184,0.002773555,0.0003423254],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005237381,0.0003841612,0.00451655,0.0005402708,0.0003961092,0.0001824289,0.001092797,0.5503093,0.01227972,0.03596104,0.008617776,0.385196],"study_design_scores_gemma":[0.00001486818,0.00006406623,0.0005986238,0.00002253965,0.00001626118,0.00009990799,0.00004846495,0.9695364,0.00332942,0.02436535,0.001879956,0.00002410477],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009016815,0.0002422649,0.9887096,0.0001491658,0.00002367676,0.00007796454,0.0000719346,0.0009980975,0.0007104533],"genre_scores_gemma":[0.3132611,0.0002316572,0.6829531,0.0003462082,0.0001863792,0.0003438916,0.0008010361,0.0002520414,0.001624546],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005319247,"threshold_uncertainty_score":0.02813125,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06678280629149759,"score_gpt":0.2353020050680726,"score_spread":0.168519198776575,"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."}}