I've released Emdros preview 1.2.0.pre195. This release sees, as its perhaps most important addition, a Penn Treebank importer. This importer has been underway for a long time... I wrote the first prototype in Python about a year ago. The new implementation is written i C++, and seems to be robust. It imports the BLLIP corpus without a hiccup, as well as the TIGER corpus in Penn format. As I said in the previous post, if anyone tests it on "the real thing" (i.e., the Penn Treebank), please let me know whether it works. Thanks.
Other goodies in the new release include:
- A NOTEXIST as described here on this blog.
- Export to Annotation Graph XML format was added.
- A few bugfixes.
- Mac OS X is now a supported platform.
- A new Chunking Tool was added as an example.
- And more...
Enjoy!
Ulrik