{"id":"W7127987698","doi":"10.1109/commantel68363.2025.11368437","title":"Hybrid Machine Learning and GenAI Approach for Data Loss Prevention","year":2025,"lang":"","type":"article","venue":"","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec à Montréal","funders":"","keywords":"Paraphrase; Classifier (UML); Baseline (sea); Machine translation; Pipeline (software); Ensemble learning; Decision tree; Confusion matrix","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":[0.005502157,0.001471284,0.00172347,0.003388421,0.001076711,0.00164082,0.003260027,0.001436383,0.002032888],"category_scores_gemma":[0.006490565,0.0005828231,0.001416354,0.002482691,0.001189591,0.003026782,0.002853415,0.00291173,0.002252382],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001164434,"about_ca_system_score_gemma":0.001999761,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004328114,"about_ca_topic_score_gemma":0.006707167,"domain_scores_codex":[0.9969103,0.0008848388,0.0001516427,0.0007710778,0.001047881,0.0002342675],"domain_scores_gemma":[0.9957866,0.001468595,0.0002991457,0.0009429692,0.001374712,0.000127882],"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":[0.0005310986,0.0005841826,0.01131448,0.0002565129,0.0003386708,0.0003947957,0.0006186006,0.1258219,0.0148299,0.009765107,0.01016188,0.8253829],"study_design_scores_gemma":[0.00001667241,0.0001212496,0.0008600631,0.00001915802,0.0000449813,0.000129245,0.00007170312,0.9786125,0.007543507,0.008715165,0.003839894,0.00002581741],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02664502,0.001060181,0.9591324,0.0006838772,0.0001410592,0.0001816191,0.0003236459,0.00945536,0.002376854],"genre_scores_gemma":[0.4132234,0.0006199539,0.5700018,0.001380221,0.000287383,0.0004695407,0.002189925,0.0007213198,0.01110639],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005502157,"threshold_uncertainty_score":0.02909851,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03970518838716456,"score_gpt":0.2961216465767995,"score_spread":0.2564164581896349,"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."}}