TFClassify is a document classification algorithm implementation.
It uses word-hypervectors where each component is computed with a modified
tf-idf weighting for a group of documents.
* It is fast, faster then edit-distance or bayesian algorithms
* It is reliable and refuses to guess on data it where the certainty is too low.
* Very little data is required to start classification
:copyright: 2006 by Florian Boesch.
:license: GNU LGPL, see LICENSE for more details.