{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00007091076,0.00006067732,0.00009616506,0.00008095952,0.0001354292,0.00001873613,0.0002566766,0.00006998685,0.00001643982],"category_scores_gemma":[0.00000858515,0.00007183358,0.00003833387,0.0001614935,0.0000807676,0.0001961778,0.0001350019,0.00007219295,0.000005137341],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001772685,"about_ca_system_score_gemma":0.00001967517,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002633787,"about_ca_topic_score_gemma":0.000002536392,"domain_scores_codex":[0.9995108,0.00001545219,0.00008019628,0.0001915404,0.0001086924,0.00009336345],"domain_scores_gemma":[0.9995471,0.00004736783,0.00008573294,0.0002137002,0.00006341559,0.00004262981],"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":[0.000009468217,0.0001972077,0.01517193,0.00001991487,0.00003175971,0.0000135071,0.001055397,0.000001477988,0.003443204,0.383398,0.008542476,0.5881156],"study_design_scores_gemma":[0.001097285,0.0002390821,0.3758489,0.00006281976,0.00004155002,0.00002235195,0.0002309781,0.4239763,0.004271178,0.04934429,0.1442663,0.0005989091],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.08525475,0.00004399924,0.9099593,0.001077887,0.00002295234,0.00009645897,1.997013e-7,0.0001247643,0.003419739],"genre_scores_gemma":[0.3338748,0.00002823181,0.6634251,0.00006109337,0.000009085392,6.962986e-7,0.000001601824,0.000003336269,0.002596069],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.5875167,"threshold_uncertainty_score":0.2929288,"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."}}