<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-1402">
  <Title>CA</Title>
  <Section position="10" start_page="0" end_page="0" type="concl">
    <SectionTitle>
9 Conclusion
</SectionTitle>
    <Paragraph position="0"> CESTA is the first European Campaign dedicated to MT Evaluation. The results of the campaign will be published in a final report and be the object of an intermediary workshop between the two campaigns and a final workshop at the end of the campaign.</Paragraph>
    <Paragraph position="1"> It is a noticeable point that the CESTA campaign aims at providing a state of the art of automated metrics in order to ensure protocol reusability. The originality of the CESTA protocol lies in the combination and contrastive use of three different types of measures carried out in parallel with a Meta evaluation of the metrics.</Paragraph>
    <Paragraph position="2"> It is also important to note that CESTA aims at providing a black box evaluation of available Machine Translation technologies, rather than a comparison of systems and interfaces, that can be tuned to match a particular need. If systems had to be compared, the fact that these applications should be compared including all software lawyers and ergonomic properties, ought to be taken into consideration.</Paragraph>
    <Paragraph position="3"> Moreover apart from providing a state of the art through a Meta evaluation of the metrics used in its protocol, thanks to the setting of this original protocol that relies on the contrastive use of complementary metrics, CESTA aims at protocol reusability. One of the outputs of the campaign will be the creation of a Machine Translation evaluation toolkit that will be put at users and system developers' disposal.Acknowledgements</Paragraph>
  </Section>
class="xml-element"></Paper>