<?xml version="1.0" standalone="yes"?>
<Paper uid="H92-1061">
  <Title>Fragment Processing in the DELPHI System</Title>
  <Section position="2" start_page="0" end_page="305" type="intro">
    <SectionTitle>
1. INTRODUCTION
</SectionTitle>
    <Paragraph position="0"> We describe the fallback understanding component of the DELPHI Natural Language component of BBN's Spoken Language System. This component is invoked when DELPHI's regular chart-based unification grammar parser is unable to parse an input; it attempts to come up with a parse and semantic interpretation, or a semantic interpretation alone, based on a fragmentary analysis of the input. The fallback understanding component consists of three separate stages, which are invoked successively. First, the Fragment Generator produces a sequence of fragmentary sub-parses from the chart state left over from the unsuccessful parse. Next, two different combination modules---the Syntactic Combiner and Frame Combiner - employ alternative and complementary strategies for combining these fragments. null The Syntactic Combiner uses extended grammar rules that can sldp over intervening material to combine constituents in an attempt to re-construct a plausible parse of the input.</Paragraph>
    <Paragraph position="1"> This parse can be a clause or some other useful constituent such as an imperative VP. A semantic interpretation for this reconstructed parse is automatically provided through the  action of the grammar rules.</Paragraph>
    <Paragraph position="2"> The Frame Combiner is invoked when the Syntactic Combiner is unsuccessful. It utilizes a set of pragmatic slot-filling schemata that embody the goals that ATIS users most commonly have, such as finding a flight or fare that satisfies some set of constraints, or asking about ground transportation between an airport or a city. As such, it detennines only a semantic interpretation and not a parse.</Paragraph>
    <Paragraph position="3"> The intent of this multi-step approach to fallback processing is to provide a smoother path between the accuracy but fragility of regular parsing on the one hand, and the robustness but possible inaccuracy of schemata-based methods on the other.</Paragraph>
    <Paragraph position="4"> The remainder of the paper is taken up with detailed description of each component. The next section, Section 2, describes the Fragment Generator. Section 3 describes the Syntactic Combiner and Section 4 the Frame Combiner. Finally Section 5 gives the February 1992 NL and SLS evaluation test results for these components, separate and combined, and our conclusions based on these results.</Paragraph>
  </Section>
class="xml-element"></Paper>