Showing cs.CLShow all
2 papers · 1 filter
cs.CL2001
Classes for Fast Maximum Entropy Training
Joshua Goodman
Maximum entropy models are considered by many to be one of the most promising avenues of language modeling research. Unfortunately, long training times make maximum entropy researc…
cs.CL2001
A Bit of Progress in Language Modeling
Joshua Goodman
In the past several years, a number of different language modeling improvements over simple trigram models have been found, including caching, higher-order n-grams, skipping, inter…