<?xml version="1.0" standalone="yes"?>
<Paper uid="E93-1027">
  <Title>Linguistic Knowledge Acquisition from Parsing Failures</Title>
  <Section position="8" start_page="229" end_page="230" type="concl">
    <SectionTitle>
7 Conclusion
</SectionTitle>
    <Paragraph position="0"> We proposed in this paper a new framework which acquires linguistic knowledge from parsing failures.</Paragraph>
    <Paragraph position="1"> Linguistic knowledge acquisition been studied so far by two extreme approaches. One approach assumes very little prior knowledge and tries to induce most of linguistic knowledge from scratch, while the other assumes existence of almost complete knowledge and tries only to learn the probabilistic properties from corpora. Our approach is between these two extremes. Although it assumes existence of rather comprehensive linguistic knowledge, it tries to create new units of knowledge which deal with specificities of given sublanguages.</Paragraph>
    <Paragraph position="2"> Considering the diverse nature of sublanguages and the essential difficulties involved in inductive processes, we believe that our approach has practical advantages over the other approaches as well as interesting theoretical implications. However, the re- null ~-~l-~entence Vana es are mltla lze to te nu string. 128323 The default blocking factor is 20 blocks. 1127131111311 There is no way selectively to follow symbolic links. II 19 \[ 6 \[ 1 II 26 I When closed, clock displays a clock face. II 1 I 0 I 0 II 1 I The default is DELETE. II 0l 41 0 II 41 This support is normally invisible to the user. II 26 \[ 13 \[ 3 11 42 \[ The output device in use is not capable of backspacing. II 40 1 14 1 -3 II 5 r I As a result, the first line must not have any superscripts. II 13 I ~ I 0 II 16 I Pathnames are restricted to 128 characters. II 0 I 1 I 0 II x I They default to the standard input and the standard output. II 12 I 5 I 1 II 18 I Remove initial definitions for all predefined symbols. II 10 I 2 I 0 II 12 I Remove any definition for the symbol name. II 2 I 0 I 0 II 2 I The most recent command is retained in any case. II 82 I 11 I 5 II 98 I Such loops are detected, and cause an error message. II 1_3 I 0 I 0 II 1_3 I Components of an expression are separated by white space. II 2 I 0 I 0 II 2 I The kernel then attempts to overlay the new process with the II 8 I 5 I 0 II 13 I desired program.</Paragraph>
    <Paragraph position="3"> Table 4: Number of Hypotheses (Sentences from the UNIX manual) search of this direction has just started and quite a few problems remain to be solved. The following shows some of these problems.</Paragraph>
    <Paragraph position="4"> * Analysis Methods of Feature Disagreements: Unlike robust parsing of ill-formed input, we have to identify real causes of disagreements and create a set of sub-hypotheses on real causes. In many cases, feature disagreements are caused by lack of or improper lexical descriptions. null * Plausibility Rating of Hypotheses: As we saw in Section 6, the corpus-based component has to take into consideration several factors, such as remedial powers and specificities of individual hypotheses, relative frequencies of hypotheses (like fault rates), competing relationships among them, etc. in order to rate the plausibility of individual hypotheses. However, the observation in Section 6 is still very sketchy.</Paragraph>
    <Paragraph position="5"> In order to design the corpus-based component, we need more detailed observation of the nature of hypotheses generated by GRHP.</Paragraph>
    <Paragraph position="6"> * Further Restrictions on Viable Hypotheses: Although the current criteria of redundant hypotheses reduce significantly the number of hypotheses, there still remain cases where more than thirty hypotheses are generated.</Paragraph>
    <Paragraph position="7"> * Refinement of Generated Hypotheses: The current version of GRHP only generates structural skeletons of new rules. These structural skeletons should be accompanied by conditions on features. In particular, it would be crucial in practical applications for GRHP to generate hypotheses of lexical descriptions with fuller feature specifications.</Paragraph>
  </Section>
class="xml-element"></Paper>