<?xml version="1.0" standalone="yes"?>
<Paper uid="P84-1009">
  <Title>APPLICATIONS OF A LEXICOGRAPHICAL DATA BASE FOR GERMAN</Title>
  <Section position="6" start_page="35" end_page="36" type="concl">
    <SectionTitle>
IV RESULT BANK
</SectionTitle>
    <Paragraph position="0"> Whereas text bank and dictionary bank supply the lexicographer with linguistic information, the result bank will be empty at the beginning of a project; it consists of a set of forms which are the frames for the word articles. Into these forms the lexicographer enters the (often preliminary) results of his work, which will be altered, amended or shortened and interrelated with other word articles (e.g. via synonymy or antonymy) in the course of compilation; he copies into those forms relevant textual samples from the text bank and useful information units from the dictionary bank.</Paragraph>
    <Paragraph position="1"> Access via terminal is not only possible to any file representing a word article but also to any record representing a category of explication.</Paragraph>
    <Paragraph position="2"> The result bank, which can be constructed within the framework of any standard data base management system, thus permits consultation and comparison on any level of lexical description. Descriptive uniformity in the morphosyntactical categories seems easy enough. But as has been shown in a number of studies, e.g. by Mugdan (1984), most existing dictionaries  abound in discrepancies and inaccuracies which easily can be avoided by cross-checking within the result bank.</Paragraph>
    <Paragraph position="3"> More difficult is homogeneity in the semantic description of the vocabulary, representing a partly hierarchical, ~artly associative net of conceptual relations. The words used in semantic explications must be used only in the same sense or senses in which they are defined under their respective heard words. These tasks can be carried out easier within a data base system. Furthermore, the result bank will support collecting and comparing the related elements of groups such us:  - all verbs with the same sentence patterns - all adjectives used predicatively only - all nouns denoting tools - all words rated as obsolete - the vocabulary of automobile  engineering.</Paragraph>
    <Paragraph position="4"> Files will differ from word class to word class, as particles or adverbs cannot be describend within the same cluster of categories as nouns or verbs. Similarily, macrostructure and microstructure will not be the same for any two dictionaries.</Paragraph>
    <Paragraph position="5"> Still Categories should be defined in such a way that the final version of the dictionary can be copied into the dictionary bank without additional manual work.</Paragraph>
    <Paragraph position="6"> After the dictionary has been compiled, it can be used as copy, using standard editing programs to produce the printed version directly from the result bank. At that level, strict formatting is no longer necessary and should be abandoned, whereever possible, in favour to economy of space.</Paragraph>
    <Paragraph position="7"> Work on the result bank will begin in autumn 1984. The pilot version of it will be applied to the current main dictionary project of the Institute, i. e. the &amp;quot;Manual of Hard Words&amp;quot;, which at present is still in its planning stage. Even in its initial version, however, LEDA will be accessible and applicable for other lexicographical projects as well.</Paragraph>
  </Section>
class="xml-element"></Paper>