Since RTextTool’s unveiling at the 2011 Cap Conference in Catania
, the development team has been busy working on refinements to the package. This includes a number of changes to simplify the API, improve analytics, decrease memory use, and increase functionality. We’ve added support for another low-memory algorithm (GLMNET) in addition to the two existing low-memory algorithms (SVM and MAXENT). Additionally, we’ve added n-fold cross validation for all nine algorithms.
Last, but certainly not least, we’ve fully documented RTextTools and are in the process of publishing the first paper about the package. Although RTextTools v1.1 is not yet released, you can get a sneak-peek of the functionality by glancing at the documentation.