{"id":"W1958345963","doi":"10.1007/978-3-540-88192-6_26","title":"Using Data Mining Methods to Predict Personally Identifiable Information in Emails","year":2008,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Calgary; Research and Productivity Council","funders":"","keywords":"Computer science; Private information retrieval; Information extraction; Data mining; Quality (philosophy); Information retrieval; Data science; Computer security","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":[0.002062799,0.0008820363,0.0009137,0.005731312,0.0006744729,0.002015011,0.0009681494,0.001427128,0.0006732217],"category_scores_gemma":[0.01033395,0.000359094,0.001037433,0.003256297,0.0003579202,0.002104466,0.0006214663,0.001363002,0.001176016],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005066963,"about_ca_system_score_gemma":0.0005417897,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001817375,"about_ca_topic_score_gemma":0.002257185,"domain_scores_codex":[0.9982452,0.0004017422,0.00024839,0.0004022291,0.0005871983,0.0001153503],"domain_scores_gemma":[0.9900066,0.007234944,0.0008174652,0.0006045304,0.001130719,0.0002057422],"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":[0.0006154115,0.001824196,0.2418488,0.0004324697,0.0004974668,0.0005975366,0.0005302021,0.04720777,0.0100037,0.002764618,0.007852535,0.6858254],"study_design_scores_gemma":[0.00002839178,0.0003252684,0.03448892,0.00009746965,0.0002090475,0.0007016979,0.0003452929,0.9345174,0.01463938,0.01101977,0.003571779,0.00005558836],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6588563,0.003034983,0.3223248,0.001329957,0.0003879463,0.0004777191,0.005931886,0.002512838,0.00514359],"genre_scores_gemma":[0.8220407,0.0009990818,0.1678723,0.0001402261,0.000285367,0.0001871874,0.005818874,0.00006288118,0.002593451],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005731312,"threshold_uncertainty_score":0.01090926,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07542289935797529,"score_gpt":0.3271868826411642,"score_spread":0.2517639832831889,"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."}}