{"id":"W2084585486","doi":"10.1016/j.eswa.2012.01.210","title":"Feature evaluation for web crawler detection with data mining techniques","year":2012,"lang":"en","type":"article","venue":"Expert Systems with Applications","topic":"Advanced Malware Detection Techniques","field":"Computer Science","cited_by":77,"is_retracted":false,"has_abstract":false,"ca_institutions":"York University","funders":"","keywords":"Computer science; Web crawler; Denial-of-service attack; Session (web analytics); Web server; Precision and recall; The Internet; Web mining; Data mining; World Wide Web; Web page; Information retrieval","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":[0.0006898039,0.000204019,0.0001807628,0.0001679232,0.0003640686,0.0001367252,0.0007993557,0.0001157988,0.000001687537],"category_scores_gemma":[0.00003231822,0.0001567065,0.00002213744,0.0005943921,0.00004398075,0.001545959,0.0001292893,0.0001104897,0.000006584202],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001667808,"about_ca_system_score_gemma":0.00009759902,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002025119,"about_ca_topic_score_gemma":0.00003206961,"domain_scores_codex":[0.9984329,0.00006802756,0.0001937868,0.0005736157,0.0004116447,0.0003200569],"domain_scores_gemma":[0.9973044,0.0001076048,0.0002456221,0.001791488,0.0004383277,0.0001125876],"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":[0.0001272223,0.0004039611,0.0006630326,0.0002005673,0.0001748,8.456802e-7,0.002031749,0.0001322647,0.1157098,0.01142527,0.02892585,0.8402047],"study_design_scores_gemma":[0.0004766911,0.00028356,0.0001633891,0.0001201278,0.00004114717,0.0002076248,0.0004301536,0.05571413,0.08968206,0.000115874,0.8522063,0.0005588765],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0001654272,0.0008422597,0.9927981,0.0003096435,0.0001185499,0.003746712,0.0000204499,0.001436595,0.0005622244],"genre_scores_gemma":[0.4681532,0.00001008322,0.5127459,0.00007715572,0.0004185775,0.0183463,0.00005694829,0.00003355624,0.0001582756],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8396458,"threshold_uncertainty_score":0.6390306,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04390771811223085,"score_gpt":0.3336468362735624,"score_spread":0.2897391181613316,"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."}}