The OntoNotes Project: Building a Large Corpus of Semantically Annotated Text

Many natural language processing (NLP) applications could benefit from a richer model of text meaning than the bag-of-words and n-gram models that currently predominate. Despite theoretical interest s