{"id":"W1963709204","doi":"10.1145/1390334.1390482","title":"Semi-supervised spam filtering","year":2008,"lang":"en","type":"article","venue":"","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Spambot; Artificial intelligence; Forum spam; Spamming; World Wide Web; The Internet","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":[0.01017269,0.0007263584,0.001656688,0.002407092,0.001109046,0.001852115,0.001465943,0.001719409,0.001898807],"category_scores_gemma":[0.02483237,0.000364347,0.001043235,0.001279864,0.001678006,0.002736544,0.001414982,0.001441777,0.001989695],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001011472,"about_ca_system_score_gemma":0.001320183,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00138832,"about_ca_topic_score_gemma":0.00204901,"domain_scores_codex":[0.9911219,0.00440455,0.000566337,0.001081048,0.002464583,0.0003615565],"domain_scores_gemma":[0.9576209,0.02009022,0.002563307,0.009232618,0.009998379,0.000494594],"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":[0.001394533,0.00114263,0.03230052,0.0007624854,0.0006765029,0.0003267885,0.00104827,0.2041491,0.02013739,0.02484794,0.02633265,0.6868812],"study_design_scores_gemma":[0.00004547966,0.000182497,0.004787185,0.00005457705,0.000042516,0.000281594,0.0001099456,0.9502473,0.01669635,0.0231475,0.004360975,0.00004403259],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2049408,0.001040406,0.7773135,0.0009993468,0.0001851697,0.0002891776,0.001068929,0.004510685,0.009651908],"genre_scores_gemma":[0.799329,0.0002165391,0.1930166,0.0004183573,0.0001894479,0.0002084648,0.002746851,0.0001810626,0.003693633],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01017269,"threshold_uncertainty_score":0.05379891,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02946477153280146,"score_gpt":0.2110597670854167,"score_spread":0.1815949955526152,"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."}}