Dave Long writes:
Cool. It shouldn't require that much redundancy* to guess that
information; It looks like an order-4 markov chain (by letters)
produces English-looking output, and an order-2 one (by words) is
best for parody, as word-order-3 results in long source text
runs. [0]
So you
Dave Long writes:
Cool. It shouldn't require that much redundancy* to guess that
information; It looks like an order-4 markov chain (by letters)
produces English-looking output, and an order-2 one (by words) is
best for parody, as word-order-3 results in long source text
runs. [0]
So you
... using LZ compression to identify the language,
authorship, and topic of documents.
Cool. It shouldn't require that much
redundancy* to guess that information;
It looks like an order-4 markov chain
(by letters) produces English-looking
output, and an order-2 one (by