Currently, the rabbit parser uses annotations to store the linguistic info when parsing Rabbit sentences. This can now be improved by using the work from LexInfo. This would make the parsing and finding of OWL entities easier and less dependent on the current GATE implementation.