The PPM data compression scheme has set the performance standard in lo
ssless compression of text throughout the past decade, PPM is a finite
context statistical modelling technique that can be viewed as blendin
g together several fixed-order context models to predict the next char
acter in the input sequence, This paper gives a brief introduction to
PPM, and describes a variant of the algorithm, called PPM, which expl
oits contexts of unbounded length. Although requiring considerably gre
ater computational resources (in both time and space), this reliably a
chieves compression superior to the benchmark PPMC version, Its major
contribution is that it shows that the full. information available by
considering all substrings of the input string can be used effectively
to generate high-quality predictions, Hence, it provides a useful too
l for exploring the bounds of compression.