Following up on my comment on #18 Jumandic parses this as two separate words too :-/ but again, if it's any conslation, JDepP (trained with IPADIC or Jumandic) assigns both results to the same bunsetsu. Is JDepP a tool that might fall into the purview of a project like Ve?
JDepP looks fantastic! It definitely falls under the purview of Ve, together with similar tools like Cabocha(*). I just haven't had time to implement any other layer besides word parsing yet.
Following up on my comment on #18 Jumandic parses this as two separate words too :-/ but again, if it's any conslation, JDepP (trained with IPADIC or Jumandic) assigns both results to the same bunsetsu. Is JDepP a tool that might fall into the purview of a project like Ve?