{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002437514,0.0009852744,0.001301826,0.005685112,0.0005371025,0.001199261,0.0008081279,0.0008152894,0.0009972811],"category_scores_gemma":[0.009734444,0.0002236593,0.001021301,0.00275546,0.0002137379,0.00128493,0.0005511632,0.0005117201,0.0004836888],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005768041,"about_ca_system_score_gemma":0.0007257419,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002998595,"about_ca_topic_score_gemma":0.00373355,"domain_scores_codex":[0.9976777,0.0004503905,0.0003129368,0.0003044136,0.001060368,0.0001941764],"domain_scores_gemma":[0.9921811,0.003845051,0.0005831404,0.0007224023,0.00245385,0.0002144622],"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":[0.001296878,0.001032677,0.0733178,0.0004995628,0.0004026423,0.0002829072,0.000108372,0.02691959,0.04886375,0.0008000787,0.005706525,0.8407692],"study_design_scores_gemma":[0.00007154684,0.000898178,0.04436389,0.00004375715,0.0002857295,0.0006158128,0.0001182056,0.894799,0.0553325,0.001195611,0.002229584,0.00004626002],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7490863,0.001688738,0.2374235,0.0002010862,0.0001227758,0.0003400467,0.00237711,0.006436517,0.002323985],"genre_scores_gemma":[0.8637454,0.0002773387,0.1309259,0.00003120182,0.00004659875,0.0001709392,0.003327265,0.0001424961,0.001332769],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005685112,"threshold_uncertainty_score":0.01289099,"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."}}