Lingua::EN::Tagger 

This module uses part-of-speech statistics from the Penn Treebank
to assign POS tags to English text.   The tagger applies a bigram (two-word)
Hidden Markov Model to guess the appropriate POS tag for a word.  That means
that the tagger will try to assign a POS tag based on the available POS tags
for the word, and the POS tag assigned to its predecessor.

The tagger tends to assume unknown words are nouns, but this behavior is 
configurable.

The POS tagger can also be used to find maximal noun phrases in tagged text.  
You can run this recursively to find all possible noun phrases.

When first run, this module will try to download POS statistical data and store
it for later use in the user's home directory.