<?xml version="1.0" standalone="yes"?>
<Paper uid="P00-1065">
  <Title>Automatic Labeling of Semantic Roles</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
2 Semantic Roles
</SectionTitle>
    <Paragraph position="0"> Historically, twotypes of semantic roles have been studied: abstract roles such as Agent and Patient, and roles speci#0Cc to individual verbs such as Eater and Eaten for #5Ceat&amp;quot;.</Paragraph>
    <Paragraph position="1"> The FrameNet project proposes roles at an intermediate level, that of the semantic frame.</Paragraph>
    <Paragraph position="2"> Frames are de#0Cned as schematic representations of situations involving various participants, props, and other conceptual roles #28Fillmore, 1976#29. For example, the frame #5Cconversation&amp;quot;, shown in Figure 1, is invoked bythe semantically related verbs #5Cargue&amp;quot;, #5Cbanter&amp;quot;, #5Cdebate&amp;quot;, #5Cconverse&amp;quot;, and #5Cgossip&amp;quot; as well as the nouns #5Cargument&amp;quot;, #5Cdispute&amp;quot;, #5Cdiscussion&amp;quot; and #5Cti#0B&amp;quot;. The roles de#0Cned for this frame, and shared by all its lexical entries, include Protagonist1 and Protagonist2 or simply Protagonists for the participants in the conversation, as well as Medium, and Topic. Example sentences are shown in Table 1. De#0Cning semantic roles at the frame level avoids some of the di#0Eculties of attempting to #0Cnd a small set of universal, abstract thematic roles, or case roles such as Agent, Patient, etc #28as in, among many others, #28Fillmore, 1968#29 #28Jackendo#0B, 1972#29#29. Abstract thematic roles can be thought of as being frame elements de#0Cned in abstract frames suchas#5Caction&amp;quot; and #5Cmotion&amp;quot; which are at the top of in inheritance hierarchy of semantic frames #28Fillmore and Baker, 2000#29.</Paragraph>
    <Paragraph position="3"> The preliminary version of the FrameNet corpus used for our experiments contained 67 frames from 12 general semantic domains chosen for annotation. Examples of domains #28see Figure 1#29 include #5Cmotion&amp;quot;, #5Ccognition&amp;quot; and #5Ccommunication&amp;quot;. Within these frames, examples of a total of 1462 distinct lexical predicates, ortarget words,were annotated: 927 verbs, 339 nouns, and 175 adjectives. There are a total of 49,013 annotated sentences, and 99,232 annotated frame elements #28which do not include the target words themselves#29.</Paragraph>
  </Section>
class="xml-element"></Paper>