{"id":"W3195417950","doi":"10.1016/j.iot.2021.100444","title":"Automated detection of unstructured context-dependent sensitive information using deep learning","year":2021,"lang":"en","type":"article","venue":"Internet of Things","topic":"Digital and Cyber Forensics","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":false,"ca_institutions":"National Research Council Canada; University of Windsor; University of Victoria","funders":"","keywords":"Context (archaeology); Computer science; Unstructured data; Deep learning; Artificial intelligence; Data science; Machine learning; Information retrieval; Data mining; Big data; Archaeology; Geography","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":[0.0004095887,0.000784046,0.0005907659,0.001446444,0.0003421244,0.0006937297,0.0007608993,0.0008233987,0.0007150014],"category_scores_gemma":[0.001403567,0.0002799263,0.0004273486,0.0007763706,0.0003494803,0.001099421,0.001280744,0.001165014,0.0005942145],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003686575,"about_ca_system_score_gemma":0.0006769371,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00140669,"about_ca_topic_score_gemma":0.003068986,"domain_scores_codex":[0.9995778,0.00005280431,0.00001794367,0.00008654303,0.0001566006,0.0001082868],"domain_scores_gemma":[0.9993466,0.0002320047,0.0001111725,0.000119638,0.0001422087,0.00004840484],"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":[0.0006296156,0.0008125476,0.01999369,0.0002400579,0.0001641804,0.001133184,0.0001865551,0.07924137,0.1614547,0.006595463,0.0100532,0.7194955],"study_design_scores_gemma":[0.000005725436,0.00007590907,0.00399765,0.00002096647,0.00002490514,0.0002827851,0.00005109536,0.9596449,0.02893228,0.005359471,0.001589081,0.00001514165],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2824232,0.001285757,0.7084239,0.0006180403,0.0002217788,0.00009109719,0.0007188682,0.002333042,0.003884292],"genre_scores_gemma":[0.9103621,0.0003881676,0.08497326,0.0001857508,0.0000951643,0.00002994649,0.0008377195,0.00005223991,0.00307567],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.001446444,"threshold_uncertainty_score":0.002796948,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006587824688594163,"score_gpt":0.2080202807576186,"score_spread":0.2014324560690244,"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."}}