{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007060441,0.0001189208,0.0001194273,0.0001284451,0.0002995307,0.0001791144,0.000661343,0.00006733581,0.000002292123],"category_scores_gemma":[0.000001382016,0.0001223345,0.00008620855,0.0004433861,0.00001623886,0.000181687,0.0001420712,0.0001045167,0.00004139824],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004006228,"about_ca_system_score_gemma":0.00003061947,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001130314,"about_ca_topic_score_gemma":0.00001849696,"domain_scores_codex":[0.9989561,0.00002275375,0.0002428274,0.0004003252,0.000100102,0.0002779144],"domain_scores_gemma":[0.9989678,0.0003718663,0.00009755922,0.0003748159,0.0001073999,0.00008056517],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000002529987,0.0000620195,0.00009459999,0.00001915803,0.00001518243,4.579866e-7,0.0003128268,0.00005492975,0.05787291,0.03919952,0.004849672,0.8975162],"study_design_scores_gemma":[0.0003400382,0.00004320488,0.0007087898,0.000008902149,0.000008837094,0.00003063693,0.00001350984,0.07206836,0.03762282,0.008048315,0.8808938,0.0002127904],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0005042061,0.0000746795,0.9965674,0.0008975762,0.00005888516,0.0006894538,0.000004133643,0.0003904652,0.0008131904],"genre_scores_gemma":[0.01908952,0.000003552629,0.9794015,0.0007271314,0.0003642844,0.0002652397,0.000008650165,0.00001255026,0.0001275905],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8973034,"threshold_uncertainty_score":0.4988657,"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."}}