{"id":"W4411010477","doi":"10.1145/3714393.3726493","title":"Evaluating Website Data Leaks through Spam Collection on Honeypots","year":2024,"lang":"en","type":"article","venue":"","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Honeypot; Computer science; Botnet; Computer security; World Wide Web; The Internet","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":[0.0006055724,0.00009989942,0.00007842851,0.00008897528,0.0001983303,0.0006167513,0.0006588217,0.00005622366,0.00008981201],"category_scores_gemma":[0.00009671845,0.00008493452,0.00003032799,0.0007750877,0.00001230829,0.001265938,0.0002864892,0.0001791509,0.000727676],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006525731,"about_ca_system_score_gemma":0.00007626611,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002101684,"about_ca_topic_score_gemma":0.00006585414,"domain_scores_codex":[0.9987047,0.00006969909,0.000137898,0.0005765536,0.0003448368,0.0001662669],"domain_scores_gemma":[0.9988909,0.0001658723,0.00002476943,0.0008529055,0.00003276015,0.00003272777],"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":[0.00004564166,0.000141485,0.0001787827,0.0001148229,0.0001230715,0.00005503785,0.003544415,0.002586244,0.01672071,0.1077803,0.3419689,0.5267406],"study_design_scores_gemma":[0.0001173844,0.0002439633,0.0003421777,0.00006226595,0.000009557898,0.0000223962,0.00001478327,0.9529908,0.004650065,0.006376389,0.03502441,0.0001457619],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01856089,0.0003069741,0.9191811,0.003506132,0.00555427,0.0002184721,0.000004187737,0.001447413,0.05122055],"genre_scores_gemma":[0.9265791,0.00003270893,0.06199755,0.0008756485,0.0006478655,0.00001377648,0.00001820428,0.00001951175,0.009815644],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9504046,"threshold_uncertainty_score":0.9353046,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.219541362242104,"score_gpt":0.4062529624929193,"score_spread":0.1867116002508153,"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."}}