Reversible Data Transforms That Improve Effectiveness of Universal Lossless Data Compression Dedykowane Moim Rodzicom, Mirosławie I Januszowi Skibińskim Contents

The subject of this dissertation are universal lossless data compression algorithms. It is a very important field of research as the data compression allows to reduce the amount of space needed to store data or to reduce the amount of time needed to transmit data. This dissertation presents lossless data compression algorithms as well as most of well-known nowadays reversible data transforms that improve effectiveness of lossless data compression algorithms. The main contribution of this dissertation are two word-based textual preprocessing algorithms, which significantly improve the compression effectiveness of universal lossless data compression schemes on textual files. These algorithms have very high encoding and decoding speed, which is amortized by a better compression effectiveness. The computational complexity remains the same as these algorithms work in a linear time. Moreover, in practice they require less than 10 MB of memory.