Classification and recovery of data from flash memory

Abstract

Classification and recovery of data are critical components of digital forensics. Numerous forensic activities need the decoding of memory dumps, the interpretation of residual data, and the recovery of data from crashed hard drives. Typically, a forensic analyst is tasked with determining the underlying file types when supplied with only file fragments and no access to the file system. The motivation is to apply various methods of feature extraction from file fragments in conjunction with machine learning and artificial intelligence for the task of file format categorization. This project will evaluate, test, and compare the accuracy and performance of the Fragments-Expert software in file fragment classification.