{"id":"W2291754627","doi":"10.11575/prism/426","title":"Web content outlier mining: motivation, framework, and algorithms","year":2006,"lang":"en","type":"article","venue":"PRISM (University of Calgary)","topic":"Anomaly Detection Techniques and Applications","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"","keywords":"Outlier; Computer science; Data mining; Web mining; Information retrieval; Web page; Artificial intelligence; 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":[0.005597896,0.001875528,0.002547101,0.007941461,0.001415944,0.007099083,0.004150369,0.003256104,0.0007654162],"category_scores_gemma":[0.0163615,0.0009485359,0.002092302,0.01409078,0.003281174,0.009185622,0.004003334,0.004344362,0.00119121],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002835755,"about_ca_system_score_gemma":0.003531648,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006205387,"about_ca_topic_score_gemma":0.003349317,"domain_scores_codex":[0.9928208,0.001656239,0.0006321127,0.001607639,0.002894425,0.0003887026],"domain_scores_gemma":[0.991771,0.003901375,0.00106856,0.0009729133,0.001934013,0.0003522758],"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":[0.0002558614,0.0005373029,0.01864741,0.001667429,0.0003376398,0.0007291169,0.0009944703,0.114603,0.00507382,0.2456567,0.02650437,0.584993],"study_design_scores_gemma":[0.00003244346,0.0001267247,0.00314067,0.0003584676,0.00007892363,0.0007955331,0.0006915735,0.710568,0.004631609,0.2495418,0.02992599,0.0001082492],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.004585424,0.004616489,0.9849666,0.002084685,0.0001339775,0.0003051303,0.0005845121,0.001200989,0.001522086],"genre_scores_gemma":[0.1117665,0.009211121,0.8704754,0.0006455877,0.001073526,0.000851547,0.003260203,0.000289194,0.002426984],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007941461,"threshold_uncertainty_score":0.02960485,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01777883724610908,"score_gpt":0.1977089406328718,"score_spread":0.1799301033867628,"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."}}