{"id":"W2798768357","doi":"10.48550/arxiv.1805.00904","title":"SynTF: Synthetic and Differentially Private Term Frequency Vectors for Privacy-Preserving Text Mining","year":2018,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Term (time); Computer science; Internet privacy; Computer security; Data mining; Business; Physics","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":[0.004610489,0.0006434329,0.0007470253,0.001558367,0.0006771997,0.001452526,0.001297451,0.001421775,0.001561781],"category_scores_gemma":[0.02781166,0.0002876316,0.0006715632,0.001779251,0.001135482,0.002623505,0.001576114,0.001675425,0.001371708],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009143184,"about_ca_system_score_gemma":0.001228893,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007595284,"about_ca_topic_score_gemma":0.0006228889,"domain_scores_codex":[0.9957416,0.001704239,0.0003595888,0.0006456824,0.00135039,0.0001986355],"domain_scores_gemma":[0.9878135,0.00621068,0.001032284,0.00389972,0.0008201359,0.0002236814],"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":[0.001516157,0.0005341254,0.006978999,0.0003757447,0.0001460049,0.0004168736,0.000520355,0.2780091,0.0318638,0.08845703,0.01756071,0.573621],"study_design_scores_gemma":[0.00005854346,0.000128268,0.0007117225,0.00002245968,0.00001152692,0.0002990411,0.00005173187,0.9160653,0.01287289,0.06347875,0.006273709,0.00002605831],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04177713,0.0003826082,0.9498742,0.0006943766,0.0001301887,0.0001434508,0.0022467,0.003614073,0.001137266],"genre_scores_gemma":[0.4773462,0.0003957733,0.5101047,0.0003251185,0.0002429759,0.0004992646,0.007916309,0.0003690356,0.002800627],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004610489,"threshold_uncertainty_score":0.02438289,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05674375505681608,"score_gpt":0.1929583039478873,"score_spread":0.1362145488910713,"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."}}