<?xml version="1.0" standalone="yes"?>
<Paper uid="C92-3146">
  <Title>TOWARDS A NEW GENERATION OF TERMINOLOGICAL RESOURCES: AN EXPERIMENT IN BUILDING A TERMINOLOGICAL KNOWLEDGE BASE</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
0 INTRODUCTION
</SectionTitle>
    <Paragraph position="0"> The discipline of terminology I has received surprisingly little focussed attention in the literature of computational linguistics - an unfortunate situation given that NLP systems seem to be most successful when applied to specialized domains. We say focussed attention because when specialized lexical items are discussed in the literature, the research problems are often not clearly differentiated from the problems of non-specialized lexical items.</Paragraph>
    <Paragraph position="1"> A fundamental assumption of our research is that, while terminology can certainly benefit from advances in computational lexicology, it nonetheless has its own non-trivial research problems, which are ultimately related to the quantity and types of specialized world knowledge that terminological repositories must contain.</Paragraph>
    <Paragraph position="2"> At the Artificial Intelligence Laboratory of the University of Ottawa, we .are constructing a new type of terminological repository, COGNITERM, which is essentially a hybrid between a term bank and a knowledge base, or a terminological knowl- edge base (TKB). COGNITERM is a bilingual (French/English) TKB constructed using a generic knowledge engineering tool (CODE) that has been used in terminology, software engineering and database design applications. The COGNITERM Project (1991-94) is focussing on the domain of optical storage technologies (e.g. optical discs, drives, processes, etc.).</Paragraph>
    <Paragraph position="3"> In Section 1 of the paper, we position our research in relation to recent developments in com1 Slmce constraints preclude even a brief description of the discipline of terminology. Cf. Sager 1990.</Paragraph>
    <Paragraph position="4"> putational lexicology and knowledge engineering; in Section 2, we describe the structure of COGN1TERM as well as some of its advantages over conventional term banks; in Section 3, we outline some methodological issues that have emerged from our work.</Paragraph>
  </Section>
class="xml-element"></Paper>