Application of Cascading Rough Set-Based Classifiers on Authorship Attribution

Author attribution aims to detect the author of given text by using various features coming from texts and authors. In this study, rough set-based classifiers are employed in a cascading style for achieving author attribution. The aim of cascading is to produce homogeneous text subsets from a heterogeneous corpus and to apply author attribution onto the smaller sets. Experiments to discover the effects of cascading on author attribution are carried out and promising results are obtained.