<?xml version="1.0" standalone="yes"?>
<Paper uid="P04-1043">
  <Title>A Study on Convolution Kernels for Shallow Semantic Parsing</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Several linguistic theories, e.g. (Jackendo , 1990) claim that semantic information in natural language texts is connected to syntactic structures. Hence, to deal with natural language semantics, the learning algorithm should be able to represent and process structured data. The classical solution adopted for such tasks is to convert syntax structures into at feature representations which are suitable for a given learning model. The main drawback is that structures may not be properly represented by at features.</Paragraph>
    <Paragraph position="1"> In particular, these problems a ect the processing of predicate argument structures annotated in PropBank (Kingsbury and Palmer, 2002) or FrameNet (Fillmore, 1982). Figure 1 shows an example of a predicate annotation in PropBank for the sentence: &amp;quot;Paul gives a lecture in Rome&amp;quot;. A predicate may be a verb or a noun or an adjective and most of the time Arg 0 is the logical subject, Arg 1 is the logical object and ArgM may indicate locations, as in our example.</Paragraph>
    <Paragraph position="2"> FrameNet also describes predicate/argument structures but for this purpose it uses richer semantic structures called frames. These latter are schematic representations of situations involving various participants, properties and roles in which a word may be typically used.</Paragraph>
    <Paragraph position="3"> Frame elements or semantic roles are arguments of predicates called target words. In FrameNet, the argument names are local to a particular frame.</Paragraph>
    <Paragraph position="4">  parse-tree representation.</Paragraph>
    <Paragraph position="5"> Several machine learning approaches for argument identi cation and classi cation have been developed (Gildea and Jurasfky, 2002; Gildea and Palmer, 2002; Surdeanu et al., 2003; Hacioglu et al., 2003). Their common characteristic is the adoption of feature spaces that model predicate-argument structures in a at representation. On the contrary, convolution kernels aim to capture structural information in term of sub-structures, providing a viable alternative to at features.</Paragraph>
    <Paragraph position="6"> In this paper, we select portions of syntactic trees, which include predicate/argument salient sub-structures, to de ne convolution kernels for the task of predicate argument classi cation. In particular, our kernels aim to (a) represent the relation between predicate and one of its arguments and (b) to capture the overall argument structure of the target predicate. Additionally, we de ne novel kernels as combinations of the above two with the polynomial kernel of standard at features.</Paragraph>
    <Paragraph position="7"> Experiments on Support Vector Machines using the above kernels show an improvement of the state-of-the-art for PropBank argument classi cation. On the contrary, FrameNet semantic parsing seems to not take advantage of the structural information provided by our kernels. null The remainder of this paper is organized as follows: Section 2 de nes the Predicate Argument Extraction problem and the standard solution to solve it. In Section 3 we present our kernels whereas in Section 4 we show comparative results among SVMs using standard features and the proposed kernels. Finally, Section</Paragraph>
  </Section>
class="xml-element"></Paper>