{"id":"W7162329119","doi":"10.63282/3050-922x.ijeret-v4i2p117","title":"A Scalable Architecture for Automated Data Classification and Sensitive Information Discovery Using Artificial Intelligence","year":2023,"lang":"","type":"article","venue":"International Journal of Emerging Research in Engineering and Technology","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute on Governance","funders":"","keywords":"Scalability; Data governance; Flexibility (engineering); Knowledge extraction; Data architecture; Cloud computing; Architecture; Data classification; Data warehouse","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":[0.001293271,0.000573676,0.0006258069,0.001482421,0.0007630988,0.001898316,0.002742628,0.0008317608,0.002119748],"category_scores_gemma":[0.002117818,0.0004013695,0.0007516261,0.001842956,0.0009083551,0.003267646,0.002210499,0.001468773,0.001325061],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001286485,"about_ca_system_score_gemma":0.002639035,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008025505,"about_ca_topic_score_gemma":0.006261097,"domain_scores_codex":[0.9987838,0.0001278831,0.0001151535,0.0003583709,0.0004912047,0.0001236237],"domain_scores_gemma":[0.9985939,0.0002528381,0.0001192731,0.0004766169,0.0004296989,0.0001275925],"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":[0.0008183577,0.001001328,0.008138075,0.0004233053,0.000317682,0.0009475817,0.00064616,0.105921,0.06337828,0.05447432,0.04365939,0.7202744],"study_design_scores_gemma":[0.0000444665,0.0001260561,0.001364398,0.00003444271,0.00008362163,0.0002263385,0.0000781121,0.9344678,0.01899746,0.02592956,0.01858895,0.00005880139],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02016253,0.0003058759,0.9406708,0.0006104825,0.0001005313,0.0003012877,0.0003730173,0.0332297,0.004245802],"genre_scores_gemma":[0.4393795,0.000447892,0.5483448,0.000781747,0.00008981568,0.0004378805,0.003677063,0.0004389693,0.006402282],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008025505,"threshold_uncertainty_score":0.01595759,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2884627211744191,"score_gpt":0.4907000701104051,"score_spread":0.202237348935986,"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."}}