This paper presents a study of the implementation of a discourse parsing system, where only significant features are considered. Rhetorical relations are recognized based on three types of cue phrases (the normal cue phrases, Noun-Phrase cues and Verb-Phrase cues), and different textual coherence devices. The parsing algorithm and its rule set are developed in order to create a system with high accuracy and low complexity. The data used in this system are taken from the RST Discourse Treebank of the Linguistic Data Consortium (LDC). © Springer-Verlag Berlin Heidelberg 2003.
CITATION STYLE
Le, H. T., & Abeysinghe, G. (2003). A study to improve the efficiency of a discourse parsing system. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2588, 101–114. https://doi.org/10.1007/3-540-36456-0_11
Mendeley helps you to discover research relevant for your work.