An automated thematic role labeler and generalizer for Filipino verb arguments

• Main Authors: Samson, Briane Paul V., Alcera, Bianca Pamela, Go, Ed Oswald, Gonzales, Czarina Meg, Lim, Nathalie Rose T.
• Format: text
• Published: Animo Repository 2009
• Subjects: Filipino language > Verb > Data processing; Computational linguistics; Computer Sciences; South and Southeast Asian Languages and Societies
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/70
• Institution: De La Salle University

A lexicon is an essential resource in the Natural Language Processing research. It provides the link between the terms of a language and the semantic and syntactic properties they are associated with. For the Filipino language, only bilingual and multilingual lexicons are available electronically. Generally, the only information they contain are the translations of a term from one language to another. They do not have information on thematic roles, which are the relations of verbs and their arguments. These relations are useful because they could allow systems to check whether the required arguments are present in the sentence. To augment manual entries of the thematic roles into the lexicon, automatic learning of thematic roles of verb arguments is explored. This paper presents the resources needed, the processes, and the results. © 2009 by Briane Paul Samson Bianca Pamela Alcera, Ed Oswald Go, Czarina Meg Gonzales, and Nathalie Rose Lim.