Mapping Words into Codewords on PPM

摘要

We describe a simple and efficient scheme which allows words to be managed in PPM modelling when a natural language text file is being compressed. The main idea for managing words is to assign them codes to make them easier to manipulate. A general technique is used to obtain this objective: a dictionary mapping on PPM modelling. In order to test our idea, we are implementing three prototypes: one implements the basic dictionary mapping on PPM, another implements the dictionary mapping with the separate alphabets model and the last one implements the dictionary with the spaceless words model. This technique can be applied directly or it can be combined with some word compression model. The results for files of 1 Mb. and over are better than those achieved by the character PPM which was taken as a base. The comparison between different prototypes shows that the best option is to use a word based PPM in conjunction with the spaceless word concept.