<?xml version="1.0" standalone="yes"?>
<Paper uid="C96-1085">
  <Title>Restricted Parallelism in Object-Oriented Lexical Parsing</Title>
  <Section position="5" start_page="505" end_page="506" type="relat">
    <SectionTitle>
4 Related Work
</SectionTitle>
    <Paragraph position="0"> Research on object-oriented natural language parsing actually started with the work of Small &amp; Rieger (1982) on word experts. Based on a conceptual parsing model, this approach took a radical position on full lexic',dization and communication based on a strict message protocol. Major drawbacks concerned an overstatement of the role of lexical idiosyncrasies and the lack of grammatical abstraction and formalization.</Paragraph>
    <Paragraph position="1"> Preserving the strengths of this approach (lexicalized control), but at the sane time reconciling it with current standards of lexicalized grammar specification, the PARSETALK system can be considered a unifying approach which combines procedural and declarative specifications at the grammar level in a formally disciplined way. This also distinguishes our approach from another major stream of object-oriented natural language parsing which is almost entirely concerned with implementational aspects of object-oriented programruing, e.g., Habert (1991), Lin (1993) or Yonezawa &amp; Ohsawa (1994).</Paragraph>
    <Paragraph position="2"> The reasons why we diverge from conventional parsing methodologies, e.g., chart parsing based on  Earley- or Tomita-style algorithms, are two-fold. First, at the syntactic level, any kind of chart parsing algorithm faces combinatorial problems with non-contiguous grammar specifications (accounting for discontinuous language structures) and, in particular, extra- and ungrammatical language input (cf., e.g., Magerman &amp; Weir (1992) for probabilistic and Lee et al. (1995) for symbolic heuristics to cope with that problem). Thus, under realistic conditions, these techniques loose a lot of their theoretical appeal and compete with other approaches merely on the basis of performance measurements. Second, including semantic considerations, even if we assume efficient syntactic processing for the sake of argument, the question arises how semantic interpretations can be processed in an incremental, comparably efficient way. Though experiments have been run with packing feature structures and interleaving syntactic and semantic analyses (Dowding et al., 1994), or with the intentional under-specification of logical forms (leaving scope ambiguities of quantifiers and negations underdetermined; cf., e.g., Hobbs (1983) or Reyle (1995)), no conclusive evidences have been generated so far in favor of a general method for efficient, online semantic interpretation. As we are faced, however, with the problem to work out text interpretations incrementally and within reasonable resource bounds, we opt for a methodology that constrains the amount of ambiguous structures right at the source. Hence, the incompleteness of the algorithm trades theoretical purism for feasibility of realistic NLP.</Paragraph>
  </Section>
class="xml-element"></Paper>