<?xml version="1.0" standalone="yes"?>
<Paper uid="W97-0910">
  <Title>Ivan.P.Bretan@telia.se Robert.H.Eklund@telia.se Mats.G.Wiren@telia.se</Title>
  <Section position="6" start_page="68" end_page="69" type="concl">
    <SectionTitle>
6 Conclusions and further directions
</SectionTitle>
    <Paragraph position="0"> We have demonstrated that it is practically feasible in the case of sufficiently close languages to generalize an existing grammar for one language to produce a grammar which, through the setting of a single parameter, becokes valid for either language. As well as providing major efficiency gains over writing a grammar for the second language from scratch, this  on representative text data technique means that subsequent enhancements to the grammar, in those areas where the characteristics of the two languages are equivalent, will apply automatically to both of them.</Paragraph>
    <Paragraph position="1"> We have also described an algorithm for composition of transfer rules. We have demonstrated that it can be used to automatically compose non-trivial sets of transfer rules containing on the order of thousands of rules, and shown that by small adjustments the performance can be improved to a level only slightly inferior to that of a corresponding set of hand-coded rules. Our experience is that the amount of work involved in using these methods is only a fraction of that needed to develop similar rules from scratch.</Paragraph>
  </Section>
class="xml-element"></Paper>