NafMultiWordTagger is a module that reads the terms in a NAF representation of a text and searches for multiword sequences of terms in the English WordNet. If it finds a multiword sequence, it groups the terms as a single term with the original terms as components.
The run.sh script reads from the input stream (which should be a NAF file) and writes to the outputstream a NAF file with the multiwords represented in the term layer.
NafMultiWordTagger is licensed under GNU GPL v3.
NafMultiWordTagger is compiled on Mac OS X version 10.6.8 with Java 1.6.
The source code can be downloaded through:
https://github.com/cltl/MultiWordTagger