Generating Parenthetical Constructions [Student Research Proposal]

Banik, Eva (2007). Generating Parenthetical Constructions [Student Research Proposal]. Technical Report 2007/14; Department of Computing, The Open University.

DOI: https://doi.org/10.21954/ou.ro.0001605f

Abstract

This paper is a research proposal for a dissertation in computational Linguistics, in particular Natural Language Generation. The purpose of the research is to provide a principled account of the generation of embedded constructions (called parentheticals) and to implement the results in a natural language generation system. Parenthetical constructions are frequently used in texts written in a good writing style and have an important role in text understanding (they help the reader to differentiate between more and less important information). They have been much studied in the linguistics literature but have received no attention so far in computational linguistics. While the ability to signal the relative importance of various items in a sentence is clearly an important contributor to the effectiveness of the text, existing natural language generation systems currently do not have a principled way of handling parentheticals. The aim of the research proposed here is to create a framework to model the rhetorical properties of different types of parentheticals and the contexts that license their usage. We will develop a unified natural language generation architecture which integrates syntax, semantics, rhetorical structure and document structure into a complex representation in order to give a principled account of constraints on where and when parenthetical constructions are appropriate to generate. The system uses constraint based reasoning to reduce computational complexity and rank the output texts. The expected results of this research will enable NLG systems to generate stylistically better output and give developers more control over the generation process and the user's interpretation of the generated text.

Viewing alternatives

Metrics

Public Attention

Altmetrics from Altmetric

Number of Citations

Citations from Dimensions
No digital document available to download for this item

Item Actions

Export

About