{"id":"W4404724706","doi":"10.1007/978-981-96-0573-6_17","title":"“Is this Site Legit?”: LLMs for Scam Website Detection","year":2024,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université de Montréal","funders":"","keywords":"Computer science; 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":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0009604787,0.0005811343,0.0004663153,0.001178071,0.0004428466,0.001563189,0.002119983,0.0005032592,0.00003747784],"category_scores_gemma":[0.00006947901,0.0005439743,0.0003056987,0.0009376571,0.0003394551,0.0009454667,0.000846272,0.0009308703,0.0003650542],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004215484,"about_ca_system_score_gemma":0.0002764895,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006409678,"about_ca_topic_score_gemma":0.0002257365,"domain_scores_codex":[0.995878,0.00001959265,0.0004908201,0.002033424,0.0008979217,0.0006802202],"domain_scores_gemma":[0.9975668,0.0004292436,0.0002319398,0.001320053,0.0002814101,0.0001705981],"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":[0.00001898316,0.00002108649,0.00002142749,0.0001680284,0.00003259052,0.00004065872,0.001966917,0.004411509,0.001463925,0.01281296,0.0003681423,0.9786738],"study_design_scores_gemma":[0.0002333661,0.0003313521,0.00003223453,0.0004587325,0.00002798867,0.0001168043,9.907224e-8,0.6742805,0.01262273,0.2701902,0.04090478,0.0008012014],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0001050184,0.0008095044,0.9831125,0.001122552,0.008696729,0.0005894452,0.00001802048,0.0005001141,0.005046184],"genre_scores_gemma":[0.5597467,0.0001943524,0.4074712,0.007347337,0.006499059,0.0001401408,0.00002420774,0.0002628548,0.01831413],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9778726,"threshold_uncertainty_score":0.9997012,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01671108114218429,"score_gpt":0.2437791157903867,"score_spread":0.2270680346482024,"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."}}