Identification and Restoration of LZ77 Compressed Data Using a Machine Learning Approach
Beom Kwon, Myongsik Gong, Jungwoo Huh, Sanghoon Lee · 2018
Identifying the type of a codec that used to compress data is essential in digital forensics since many trials and errors required to restore data can be reduced. Nevertheless, most compression algorithms have been configured by using several parameters whose values can be different according to each user. Therefore, in order to restore data more effectively, the values of parameters as well as the type of the codec must be identified. In this paper, we present an identification and restoration method for Lempel-Ziv-77 (LZ77) compressed data. In the proposed method, we identify whether a given data is compressed by LZ77 or not. Moreover, we estimate the values of parameters that were used for compression. Using the estimated parameters, we restore the original data from the LZ77 compressed data. The simulation results demonstrate the feasibility and effectiveness of the proposed method with a successful compression identification and parameter estimation accuracies of 100% and 84.41%.