{"id":"W4324118134","doi":"10.1109/bdcat56447.2022.00018","title":"Mining Large Data to Create a Balanced Vulnerability Detection Dataset for Embedded Linux System","year":2022,"lang":"en","type":"article","venue":"","topic":"Advanced Malware Detection Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ontario Tech University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Hacker; Computer science; Firmware; Vulnerability (computing); Computer security; Linux kernel; Embedded system; Operating system","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00174895,0.000153176,0.000206048,0.000146098,0.0005357298,0.000113155,0.001759492,0.00003331948,0.00002581328],"category_scores_gemma":[0.0002707953,0.0001584824,0.00003711126,0.0005577469,0.000009211938,0.0008212159,0.002754176,0.0001654462,0.00000846488],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002338595,"about_ca_system_score_gemma":0.00005392787,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004509015,"about_ca_topic_score_gemma":0.0001327602,"domain_scores_codex":[0.9979144,0.000203156,0.000319018,0.0009400342,0.0002857892,0.0003376252],"domain_scores_gemma":[0.9971261,0.0002212631,0.0001181616,0.002350611,0.00007964254,0.000104241],"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":[0.001505561,0.001496627,0.0003798462,0.001113902,0.0002797809,0.0001190301,0.004720538,0.01730025,0.1810707,0.07033831,0.3063622,0.4153133],"study_design_scores_gemma":[0.0009420617,0.000799587,0.0001010539,0.00002028755,0.00001555273,0.00009857373,0.0009334097,0.5890964,0.07733882,0.0007051672,0.3293709,0.0005781061],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002155423,0.00001025027,0.9913296,0.0002047715,0.0003583912,0.0009167999,0.003559059,0.001346397,0.0001193669],"genre_scores_gemma":[0.6642417,5.221887e-7,0.3329663,0.0008341554,0.00004122582,0.0009040962,0.0008813713,0.00001685113,0.0001137776],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6620863,"threshold_uncertainty_score":0.6462724,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05378122401371729,"score_gpt":0.3368215604977142,"score_spread":0.283040336483997,"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."}}