<?xml version="1.0" standalone="yes"?>
<Paper uid="H94-1026">
  <Title>Toward Multi-Engine Machine Translation</Title>
  <Section position="6" start_page="149" end_page="149" type="concl">
    <SectionTitle>
5. FUTURE WORK
</SectionTitle>
    <Paragraph position="0"> Ultimately, a multi-engine system depends on the basic quality of each particular engine. We expect the performance of some of the individual engines (especially, KBMT and EBMT) to grow. Consequently, the multi-engine environment will improve, as larger static knowledge sources are added and the scoring mechanism is further adjusted. We expect to gain insight into how to improve the scoring mechanism: we plan a battery of tests to help adjust the coefficients on the function which combines the individual scores in the final score. We plan to use a standard regression mechanism to modify these scores based on feedback from having humans select the best covers for test texts. We expect such calibration further to optimize the system to produce the best possible output from the set of available candidate translations produced by the multiple engines. We also intend to develop a method of how to empirically assess the expected output quality of each translation engine based on its available resources, such as dictionaries, glossaries, grammars, paragel corpora, etc.</Paragraph>
  </Section>
class="xml-element"></Paper>