Abstract
This paper presents a novel approach to the development of anaphoric annotation of large corpora based on the use of semantic information to help the annotation process. The anaphora annotation scheme has been developed from a multilingual point of view in order to annotate three corpora: one for Catalan, one for Basque and one for Spanish. An anaphora resolution system based on restrictions and preferences has been used to aid the manual annotation process. Together with morpho-syntactic information, the system exploits the semantic relation between the anaphora and its antecedent.
Cite
CITATION STYLE
Saiz-Noeda, M., Navarro, B., & Izquierdo, R. (2004). Semantic-aided anaphora resolution in large corpora development. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 3230, pp. 314–327). Springer Verlag. https://doi.org/10.1007/978-3-540-30228-5_28
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
