Asian Journal of Information Technology
ISSN: Online 1993-5994ISSN: Print 1682-3915
Abstract
This study presents a new algorithm for the resolution of anaphoric expressions which uses a few ideas largely accepted by the scientific community in computational linguistics as part of the resolution process. The algorithm driven by a robust morphosyntactic parser uses a set of constraints, which the candidates must not violate, as well as a set of preferences which discriminate among them. We also incorporate the notion of focused entities, and eliminate from further consideration entities included in insertions. The algorithm, written in Prolog, uses Definite Clause Grammars (DCG) for the implementation of the parser as well as the constraints and the preferences. Finally we discuss the results and the efficiency of the algorithm in light of related work in the field.
How to cite this article:
Allaoua Refoufi . A Multiple Knowledge Sources Algorithm for Anaphora Resolution.
DOI: https://doi.org/10.36478/ajit.2006.48.53
URL: https://www.makhillpublications.co/view-article/1682-3915/ajit.2006.48.53