Computational Techniques for Referring Expressions Resolution

• Grant number: 07/07356-4
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: February 01, 2008
• End date: January 31, 2009
• Field of knowledge: Physical Sciences and Mathematics - Computer Science
• Principal Investigator: Ivandre Paraboni
• Grantee: Ramon Re Moya Cuevas
• Host Institution: Escola de Artes, Ciências e Humanidades (EACH). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

The computational resolution of referring expressions such as personal pronouns (e.g., him), definite descriptions (e.g., the blue building, on the left) and others is a common task for a variety of NLP applications, ranging from text understanding to machine translation or summarization. Although it has received a great deal of attention for many years, the task of co-reference resolution remains a computational problem yet to be overcome, for which a large number of language or domain-specific solutions have been proposed. In this work we propose an investigation on machine learning approaches to pronoun resolution based on existing training data derived from a corpus developed as part of an ongoing research project, aiming to further the knowledge on Portuguese language processing and to introduce the applicant to the NLP research field and standard AI techniques in general.