{"id":"W2348504122","doi":"","title":"A Method for Combating Web Spam Pages","year":2007,"lang":"en","type":"article","venue":"Microcomputer applications","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Spamdexing; Spambot; Search engine; Classifier (UML); Web page; Information retrieval; Bag-of-words model; Data mining; World Wide Web; Forum spam; The Internet; Web search engine; Web search query; Artificial intelligence; Spamming","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001327525,0.001427332,0.001157365,0.006503952,0.00172105,0.001438037,0.001101003,0.002094234,0.003516228],"category_scores_gemma":[0.00391205,0.0004630611,0.0007590913,0.00165711,0.0009244714,0.001984141,0.0007576214,0.001078938,0.003856908],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004956701,"about_ca_system_score_gemma":0.001133848,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001328258,"about_ca_topic_score_gemma":0.0014757,"domain_scores_codex":[0.9980982,0.0003186575,0.0001019362,0.0002938036,0.001079752,0.0001076319],"domain_scores_gemma":[0.9965353,0.0009100685,0.0003557248,0.0004331974,0.001616262,0.0001495027],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0002861484,0.0004548526,0.004330005,0.0003753356,0.0001293566,0.0003682818,0.0002578718,0.003580901,0.08456003,0.007581951,0.01487574,0.8831995],"study_design_scores_gemma":[0.0003433141,0.001360139,0.01561149,0.0002490818,0.0003540801,0.007385292,0.0003359663,0.537084,0.2899952,0.02283409,0.1240497,0.0003978483],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.028238,0.001241486,0.9504723,0.0006887095,0.0004366151,0.0007731389,0.0002587458,0.01185692,0.006034023],"genre_scores_gemma":[0.1373274,0.0005829104,0.8478572,0.0004842113,0.0004020923,0.0005134356,0.0004305373,0.0003289227,0.01207323],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006503952,"threshold_uncertainty_score":0.01176298,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01661931472388831,"score_gpt":0.2970084248152411,"score_spread":0.2803891100913528,"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."}}