Automatically Classifying Documents by Ideological and Organizational Affiliation

Abstract

We show how an Arabic language religious-political document can be automatically classified according to the ideological stream and organizational affiliation that it represents. Tests show that our methods achieve near-perfect accuracy.