The Stanford parser integration should allow to use pre-existing POS tags. I
believe to do that, the data that goes to the parser must be a list of Word
objects - possibly there is more information in the JavaDoc or on the parser
mailing list.
This should be optional and turned off by default.
Original issue reported on code.google.com by richard.eckart on 29 Jan 2013 at 8:56
Original issue reported on code.google.com by
richard.eckart
on 29 Jan 2013 at 8:56