Technical

ATLAS features a module for automatic multi-label multi-class categorization of documents. The following algorithms are currently included in the module Naïve Bayesian, Relative Entropy and Class featured centroid classifier (Guan, Zhou, Guo, 2009). The conducted experiments showed that the above algorithms provide reasonable classification accuracy and are much faster than more complex methods (such as Support Vector Machines). Nevertheless, the above list is not final, and, based on the internal experiments, new classification methods and strategies to be included in later versions of the system.