Combining Bag-of-Words and Bag-of-Concepts representations for Arabic text classification

This paper introduces a set of new approaches for text representation for automatic classification of Arabic textual documents. These approaches are based on combining the well-known Bag-of-Words (BOW) and the Bag-of-Concepts (BOC) text representation schemes and utilizing Wikipedia as a knowledge base. The proposed representations are used to generate a vector space model, which in turn is fed into a classifier to categorize a collection of Arabic textual documents. Three different machine learning based classifiers have been utilized in this work. Performance of proposed text representation models is evaluated in comparison to using a standard BOW scheme and a concept-based scheme, as well as recently reported similar text representation schemes that are based on augmenting the standard BOW with the BOC.