Source encoding using syntactic information source models

Abstract

The use of syntactic information source models for the source encoding (data compression) of messages is described. Syntactic models formulated using context-free grammars augmented with derivation step probabilities are considered. Using an arithmetic coder as the low-level encoding unit, it is shown how practical encoding systems can be constructed from such models. Application of the techniques to the encoding of syntactically correct Pascal computer programs is described, and additional techniques including the use of symbol tables are introduced. The resultant syntactic encoders achieve compression of Pascal programs approaching 90%.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>