A Knowledge-Based Approach for Provisions' Categorization in Arabic Normative Texts

This paper studies the problem of automatic categorization of provisions in Arabic normative texts. We propose a knowledge-based categorization approach coupling a taxonomy of Arabic normative provisions' categories, an Arabic normative terminological base and a rule-based semantic annotator. The obtained model has been trained and tested over a collection of Arabic normative texts collected from the Official Gazette of the Republic of Tunisia. The performance of the approach was evaluated in terms of Precision, Recall and F-score in order to categorize instances over 14 normative categories. The obtained results over the test dataset are very promising. We have obtained 96.4 % for Precision, 96.06 % for Recall and 96.23 % for F-score.