<?xml version="1.0" standalone="yes"?>
<Paper uid="C94-2211">
  <Title>A Knowledge Acquisition and Management System for Morphological Dictionaries</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1. Introduction
</SectionTitle>
    <Paragraph position="0"> It has long been recognized that there is a need for reusable dictionaries for NLP-systems. A system for the development and maintenance of such dictionaries can be viewed as a knowledge acquisition and management tool. The major objectives from this perspective are support for efficient dictionary construction, and maintenance of consistency within the database.</Paragraph>
    <Paragraph position="1"> As a starting point, it makes sense to take morphological dictionaries of a type that can at least solve the mapping problem between text words and dictionary entries. This problem directly involves inflection, (graphic) clitics (e.g. German zum = zu dent), and multi-word units, while wordformation is necessary for the recognition of newly coined words. Concentrating on these morphological dictionaries recommends itself because the information they contain is relatively well-understood (compared to e.g. semantics), can be isolated fairly well from more application-specific types of knowledge, and is necessary for almost every NLP-application.</Paragraph>
    <Paragraph position="2"> In this project note, we will first consider finite-state morphology systems from the perspective of knowledge acquisition and management, and observe some shortcomings (section 2).</Paragraph>
    <Paragraph position="3"> Then Word Manager (WM) is presented, and it is shown how WM avoids these drawbacks (section 3). In section 4, an overview of the present state of the WM-project is given.</Paragraph>
  </Section>
class="xml-element"></Paper>