Towards a New Approach for Disambiguation in NLP by Multiple Criterian Decision-Aid
Youssef Hoceini, Mohamed Amine Chéragui, Moncef Abbas · The Prague Bulletin of Mathematical Linguistics · 2011
Towards a New Approach for Disambiguation in NLP by Multiple Criterian Decision-Aid The aim of this paper is to present a combination of NLP and Multiple Criteria Decision-Aid (MCDA) in order to reach an effective analysis when dealing with linguistic data from various sources. The coexistence of these two concepts has allowed us, based on a set of actions and criteria, to develop a coherent system that integrates the entire process of textual data analysis (no-voweled Arabic texts) into decision making in case of ambiguity. Our solution is based on decision theory and an MCDA approach with a TOPSIS technique. This method allows the multi-scenario classification of morphosyntactical ambiguity cases in order to come out with the best performance and reduce the number of candidate scenarios.