I just released LingPipe version 4.0.1. As usual, you can read more about and download the latest version from the
The latest version of the book is soon to follow, with much more LingPipe-specific content.
- fixes some bugs and brings definitions in line with standards for
- added more documentation and utilities to
classify.LogisticRegressionClassifierregression so that likelihood gradient updates are blocked along with prior updates,
- adds an implementation
toString()methods in all tokenizer factories to expose nested structure,
- added a collection utilities class
- added some more utility methods for utilities classes
- added utility methods for
- fixed a bug in sparse vector addition with other sparse vectors (which nothing else depended on), and
- brings the tokenized LM tutorial in line with LingPipe 4.