Year: 2013
Type: article
Abstract: In this study, we address the problem of improving variable-length-to-fixed-length codes (VF codes). A VF code is an encoding scheme that uses a fixed-length code, which provides easy access to compressed data. However, conventional VF codes generally have an inferior compression ratio compared with variable-length codes. A method proposed by Uemura et al. in 2010 delivered a good compression ratio that was comparable with that of gzip, but it was very time consuming. In this study, we propose a new VF coding method that applies a fixed-length code to a set of rules extracted using the Re-Pair algorithm, which was proposed by Larsson and Moffat in 1999. The Re-Pair algorithm is a simple offline grammar-based compression method, which has good compression-ratio performance with moderate compression speed. We also present experimental results, which demonstrates that our proposed coding method is superior to the existing VF coding method.
Source: IPSJ Online Transactions
Authors: Satoshi Yoshida, Takuya Kida
Institution: Hokkaido University
Language: English
Subfield: Theoretical Computer Science
Field: Mathematics
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W2086374114