The Brill Tagger

• An old, and still solid, piece of technology is the Brill Part of Speech tagger. It's in GATE but also freely available elsewhere.
• Part of speech tagging looks through open text and gives particular words their lexical category.
• Examples of lexical categories are verbs, nouns, adjectives and determiners.
• Typically, later stages of processing (e.g. parsing) take this as input.
• Note that most open class words are lexically ambiguous. E.g. note is both a verb and a noun.
• The Brill tagger is trainable. That is, you give it a bunch of text where the parts of speech are known.
• It then learns to tag unseen text.
• It's about 97% effective.