<?xml version="1.0" standalone="yes"?>
<Paper uid="A00-1017">
  <Title>A Representation for Complex and Evolving Data Dependencies in Generation</Title>
  <Section position="8" start_page="124" end_page="125" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> The representation scheme we have proposed here is designed specifically to support the requirements of the current state-of-the-art NLG systems, and our pilot implementation demonstrates the practical applicability of the proposal. Tangled, partial and mixed structures are of obvious utility to any system with a flexible control strategy and we have shown here how the proposed representation scheme supports them. By recording the derivational history of computations, it also supports decisions which partly depend on earlier stages of the generation process (e.g., possibly, lexical choice) and revision-based architectures which typically make use of such information. We have shown how the representation scheme might be the basis for an inter-module communication model, the whiteboard, which supports a wide range of processing strategies that require the representation of complex and evolving data dependem cies. The fact that the whiteboard is cumulative, or monotonic in a logical sense, means that the whiteboard also supports reasoning about the behaviour of NLG systems implemented in terms of it. This is something that we would like to exploit directly in the future.</Paragraph>
    <Paragraph position="1"> The reimplementation of the CGS system in the RAGS framework was a challenge to the framework because it was a system that had already been developed completely independently. Even though we did not always understand the detailed motivation for the structure of CGS being as it was, within a short time we reconstructed a working system with modules that corresponded closely to the original CGS modules. The representation scheme we have proposed here was a key ingredient in giving us the flexibility to achieve the particular processing scheme used by CGS whilst remaining faithful to the (relatively simple) RAGS data model.</Paragraph>
    <Paragraph position="2">  The representation scheme is useful in situations where modules need to be defined and implemented to work with other modules, possibly developed by different people. In such cases, the representation scheme we propose permits precise definition of the interfaces of the modules, even where they are not restricted to a single 'level' of representation. Even though the control structure of CGS is quite simple, we found that the use of a centralised whiteboard was useful in helping us to agree on interfaces and on the exact contribution that each module should be making. Ultimately, it is hoped that the use of a scheme of this type will permit much more widespread 'plug-and-play' among members of the NLG community.</Paragraph>
  </Section>
class="xml-element"></Paper>