In Comparison with the study of compressing texts that consists of 1 byte character, there are few algorithms about Japanese text compression using the tendencies of Japanese language. Only we can see some methods that are used before compressing by some usual algorithms. Since Japanese has different system from English, there is some possibility to compress shorter. This paper suggests an algorithm that uses such tendencies. This algorithm is based on the conventional algorithm Prediction by Partial Match (PPM). PPM is an algorithm that predicts next character by the past contexts. In this paper it is improved to be suitable for the Japanese texts.
展开▼