File Information
File: 05-lr/acl_arc_1_sum/cleansed_text/xml_by_section/concl/90/j90-3001_concl.xml
Size: 1,715 bytes
Last Modified: 2025-10-06 13:56:34
<?xml version="1.0" standalone="yes"?> <Paper uid="J90-3001"> <Title>RESOLVING QUASI LOGICAL FORMS</Title> <Section position="16" start_page="0" end_page="0" type="concl"> <SectionTitle> 7 FURTHER RESEARCH </SectionTitle> <Paragraph position="0"> We are currently developing our model for QLF resolution within the framework of a combined language processing and contextual reasoning system called CLARE, of which the CLE is the language component. This includes work on the treatment of ellipsis as QLF resolution. Some simple cases are already handled, such as the representation of an utterance consisting solely of a noun phrase as a form (...,P,\[P,NP\]) and resolving it by substituting the noun phrase meaning into the RQLF for a preceding sentence.</Paragraph> <Paragraph position="1"> We hope to extend this treatment to comparatives and Computational Linguistics Volume 16, Number 3, September 1990 143 Hiyan Alshawi Resolving Quasi Logical Forms other elliptical expressions whose resolutions are partially governed by syntax. Another signif.cant aspect of QLF resolution that we have not yet addressed is determining the temporal indices implicit in QLF tense and aspect operators.</Paragraph> <Paragraph position="2"> Perhaps the most challenging problem in automatic language processing, that of simulating the role of plausibility in understanding, remains the most difficult part of QLF resolution. Our framework isolates this problem into the more concrete one of assigning plausibility ratings to RQLF interpretations. We will be looking at ways of exploiting inferential and probabilistic reasoning with domain knowledge in developing solutions to this problem.</Paragraph> </Section> class="xml-element"></Paper>