<?xml version="1.0" standalone="yes"?>
<Paper uid="M91-1008">
  <Title>GTE'S TEXT INTERPRETATION AID (TIA) : MUC-3 TEST RESULTS AND ANALYSI S</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
INTRODUCTION
</SectionTitle>
    <Paragraph position="0"> GTE has actively participated in the Third Annual Message Understanding Conference (MUC-3) sponsored by the Naval Ocean System Center (NOSC) for the past 7 months using its natural language processing system, TIA (Text Interpretation Aid)[5]. During this period, TIA system development consisted of two tasks: (1) lexicon tool development, and (2) tailoring TIA to meet the specific needs of the MUC-3 domain, i .e., terrorism in Lati n American countries. Lexicon tool development facilitates system (lexicon) adaptation to new domains (terrorism, drug interdiction, third world launchings, etc.) by semi-automating the manual task of entering words/phrases into TIA's lexicons. Tailoring the existing TIA system to parse and extract MUC-3 domain messages, has allowed GT E to participate in the MUC-3 conference tasks.</Paragraph>
    <Paragraph position="1"> The purpose of this paper is to report our techniques and findings during these 7 months of MUC-TIA syste m development. Additionally, GTE's official scores for Phase I and Phase II of MUC-3 will be disclosed and discussed . The remainder of this paper is divided into seven sections describing scoring results, justification and analysis o f scores, system development effort, limiting factors, training, reusability of the resultant MUC-TIA system, and fina l conclusions.</Paragraph>
  </Section>
class="xml-element"></Paper>