<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-3122">
  <Title>Language Models and Reranking for Machine Translation</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Statistical machine translation (SMT) systems combine a number of translation models with one or more language models. Adding complex language models in the incremental process of decoding is a very challenging task. Some language models can only score sentences as a whole. Also, SMT decoders generate during the search process a very large number of partial hypotheses and query the language model/models 1.</Paragraph>
    <Paragraph position="1"> The solution to these problems is either to use multiple iterations for decoding, to make use of the complex LMs only for complete hypotheses in the search space or to generate n-best lists and to rescore the hypotheses using also the additional LMs. For  scribed in Section 3, the 3-gram LM was queried 27 million times (3 million distinct queries).</Paragraph>
    <Paragraph position="2"> the WMT 2006 shared task we opted for the reranking solution. This paper describes our solution and results.</Paragraph>
  </Section>
class="xml-element"></Paper>