by wellner
Carafe is an efficient implementation of Conditional Random Fields and related algorithms in OCaml targeted at text processing applications. Features include: Voted Perceptron and log-likelihood training, feature induction, an SGML/XML front-end, etc.
A handful of major new features have been implemented: * Fast, robust stochastic gradient descent using Periodic Stepsize Adjustment (PSA) * Disk-caching. Instantiated features for sequences can be cached on disk, allowing training over datasets ...
Simplified the build process. Works on Cygwin now.
A general framework for using non-factored features has been added to Carafe - these features don't explicitly predicate over the output variable assignments (as is typical with CRFs). This can be used for discriminative word-alignment or "sequence ...
Carafe uses ocamlbuild to build. Added functionality for non-factored features specifically aimed at tasks such as discriminative word-alignment ( Blunsom & Cohn 2006)
Copyright © 2009 SourceForge, Inc. All rights reserved. Terms of Use