<?xml version="1.0" standalone="yes"?>
<Paper uid="W03-1008">
  <Title>Identifying Semantic Roles Using Combinatory Categorial Grammar</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
2 Predicate-argument relations in
PropBank
</SectionTitle>
    <Paragraph position="0"> The Proposition Bank (Kingsbury and Palmer, 2002) provides a human-annotated corpus of semantic verb-argument relations. For each verb appearing in the corpus, a set of semantic roles is defined. Roles for each verb are simply numbered Arg0, Arg1, Arg2, etc. As an example, the entryspecific roles for the verb offer are given below:  These roles are then annotated for every instance of the verb appearing in the corpus, including the following examples:  to apply across all verbs. These secondary roles can be thought of as being adjuncts, rather than arguments, although no claims are made as to optionality or other traditional argument/adjunct tests. The secondary roles include: Location in Tokyo, outside Time last week, on Tuesday, never Manner easily, dramatically Direction south, into the wind Cause due to pressure from Washington Discourse however, also, on the other hand Extent 15%, 289 points Purpose to satisfy requirements Negation not, n't Modal can, might, should, will Adverbial (none of the above) and are represented in PropBank as &amp;quot;ArgM&amp;quot; with an additional function tag, for example ArgM-TMP for temporal. We refer to PropBank's numbered arguments as &amp;quot;core&amp;quot; arguments. Core arguments represent 75% of the total labeled roles in the PropBank data. Our system predicts all the roles, including core arguments as well as the ArgM labels and their function tags.</Paragraph>
  </Section>
class="xml-element"></Paper>