I just put up the latest draft of the LingPipe book. Since the last edition, I’ve added a chapter on character language models and added details of the various classifier interface and classification types. The location’s the same:
I’m about to roll out LingPipe 4.0.2, which patches a few minor inconsistencies that have turned up in writing the book (like process character LMs not being serializable and some remaining generic arrays that I’ve recoded with lists [and deprecated the generic array versions]).
