<?xml version="1.0" standalone="yes"?>
<Paper uid="P98-2140">
  <Title>Feature Lattices for Maximum Entropy Modelling</Title>
  <Section position="7" start_page="853" end_page="854" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> In this paper we presented a novel approach for building maximum entropy models. Our approach uses a feature collocation lattice and selects the candidate features without resorting to iterative scaling. Instead we use our own frequency redistribution algorithm. After the candidate features have been selected we, using the iterative scaling, compute a fully saturated model for the maximal constraint space and then apply relaxation to the most specific constraints.</Paragraph>
    <Paragraph position="1"> We applied the described method to several language modelling tasks such as sentence boundary disambiguation, part-of-speech tagging, stress prediction in continues speech generation, etc., and proved its feasibility for selecting and building the models with the complexity of tens of thousands constraints. We see the major achievement of our method in building compact models with only a fraction of possible features (usually there is a few hundred features) and at the same time performing at least as good as state-of-the-art: in fact, our sentence boundary disambiguater scored the highest known to the author accuracy (99.2477%) and our part-of-speech tagging model generalized for a new domain with only a tiny degradation in performance.</Paragraph>
    <Paragraph position="2"> A potential drawback of our approach is that we require to build the feature collocation lattice for the whole observed feature-space which might not be feasible for applications with hundreds of thousands of features. So one of the directions in our future work is to find efficient ways for a decomposition of the feature lattice into non-overlapping sub-lattices which then can be handled by our method. Another avenue for further improvement is to introduce  the &amp;quot;or&amp;quot; operation on the nodes of the lattice. This can provide a further generalization over the features employed by the model.</Paragraph>
  </Section>
class="xml-element"></Paper>