<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-2610">
  <Title>An Ontology-Based Approach to Disambiguation of Semantic Relations</Title>
  <Section position="6" start_page="75" end_page="76" type="concl">
    <SectionTitle>
5 Conclusion and future work
</SectionTitle>
    <Paragraph position="0"> Even though the experiments are in an early phase, the results indicate that it is possible to analyse the semantic relation a preposition denotes between two noun phrases, by using machine learning and an annotated corpus - at least within the domain covered by the ontology. Future work will therefore include annotation and investigation of a general language corpus. Also, a  morethoroughexaminationofthecorpus,morespecifically an investigation of which relations or prepositions that are most difficult to analyse. Also, we will experiment with the amount of information that we train on, not as we have already done by in- or excluding types of information, but rather the extension of the information: Could we predict the ontological type of one of the arguments by looking at the other? Finally, an explicit inclusion of the whole ontology in the learning process is on the agenda, as proposed in section 3.1.1 on page 3, in the anticipation that the learner will produce an even better model.</Paragraph>
  </Section>
class="xml-element"></Paper>