Is it possible to train mstparser on partial parse trees? If so, where can I find some information, examples, etc. about it? Do I need to modify the source code, or are there some options or ...?
I'm working on dependency parsing projection, to create training corpus for creating dependency parser for resource poor languages. The training trees created through projection are not full parse trees though, because some of the dependency links get lost during the projection process.
Thanks in advance!
Log in to post a comment.