<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-2910">
  <Title>Can Human Verb Associations Help Identify Salient Features for Semantic Verb Classification?</Title>
  <Section position="8" start_page="74" end_page="75" type="concl">
    <SectionTitle>
6 Summary
</SectionTitle>
    <Paragraph position="0"> The questions we posed in the beginning of this paper were (i) whether human associations help identify salient features to induce semantic verb classes, and (ii) whether the same types of features are salient for different types of semantic verb classes.</Paragraph>
    <Paragraph position="1"> An association-based clustering with 100 classes served as source for identifying a set of potentially salient verb features, and a comparison with standard corpus-based features determined proportions of feature overlap. Applying the standard feature choices to verbs underlying three gold standard verb classi cations showed that (a) in our experiments there is no correlation between the overlap of associations and feature types and the respective clustering results. The associations therefore did not help in the speci c choice of corpus-based features, as we had hoped. However, the assumption that window-based features do contribute to semantic verb classes this assumption came out of an analysis of the associations was con rmed: simple window-based features were not signi cantly worse (and in some cases even better) than selected grammar-based functions.</Paragraph>
    <Paragraph position="2"> This nding is interesting because window-based features have often been considered too simple for semantic similarity, as opposed to syntax-based features. (b) Several of the grammar-based nominal and adverbial features and also the window-based features outperformed feature sets in previous work, where frame-based features (plus prepositional phrases and coarse selectional preference information) were used. Surprisingly well did adverbs: they only represent a small number of verb features, but obviously this small selection can out-perform frame-based features and even some nomi- null nal features. (c) The clustering results were significantly better for the GermaNet clusterings than for the experiment-based and the FrameNet clusterings, so the chosen feature sets might be more appropriate for the synonymy-based than the situation-based classi cations.</Paragraph>
    <Paragraph position="3"> Acknowledgements Thanks to Christoph Clodo and Marty Mayberry for their system administrative help when running the cluster analyses.</Paragraph>
  </Section>
class="xml-element"></Paper>