A fast dynamic compression scheme for natural language texts

AbstractWe adapted Word-based Tagged Code (WBTC) to obtain its dynamic version. The aim of designing a dynamic version of WBTC is to adapt it for real-time transmission. The problem in the semi-static technique is to perform two passes over the source text, and therefore encoding cannot start before the whole first pass has been completed. For this reason, the semi-static technique is not applied to compress text streams. However, the adaptive or dynamic model does not have this limitation. The dynamic WBTC can be used for real-time transmission where the semi-static version is not suited at all. The experimental results show that the Dynamic WBTC is suitable for compressing text streams by utilizing low bandwidth