<?xml version="1.0" standalone="yes"?>
<Paper uid="C82-1033">
  <Title>SURFACE ANALYSIS OF QUERIES DIRECTED TOWARD A DATABASE</Title>
  <Section position="11" start_page="207" end_page="207" type="ackno">
    <SectionTitle>
4.0 RECAPITULATION
</SectionTitle>
    <Paragraph position="0"> The rapidly improving capabilities of computer hardware combined with the rapid decline in the cost of that hardware has created a situation where a major limiting factor in the utility and growth of DBMS is the inability of many people to use the complex software packages needed to access databases. One possible solution is the development of a natural language interface that can serve as an intermediary between the user and the DBMS. This would enable users to communicate with the database in their own language instead of the computers.</Paragraph>
    <Paragraph position="1"> The mechanism described maximizes surface analys~s of the query and minimizes the amount of content &amp;quot;understanding&amp;quot; needed to resolve a query. The construction of a world mechanism for each application Semantic information necessary to resolve a query is derived from existing sources such as the logical schema.</Paragraph>
    <Paragraph position="2"> The research involved in completely specifying and implementing the mechanism is directed toward two fundamental questions. They are: (I) what is the minimum amount of &amp;quot;understanding&amp;quot; of a natural language query that is needed to generate a semantically equivalent DML query and (2) how much information about a query can be derived from surface analysis.</Paragraph>
    <Paragraph position="3"> Note: A detailed comparison of this mechanism with other systems can be found in a working paper (Feinauer, 1981) on this mechanism.</Paragraph>
  </Section>
class="xml-element"></Paper>