<?xml version="1.0" standalone="yes"?>
<Paper uid="W02-0210">
  <Title>Adaptive Dialogue Systems - Interaction with Interact</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> The need for flexible interaction is apparent not only in everyday computer use, but also in various situations and services where interactive systems can diminish routine work on the part of the service provider, and also cater for the users with fast and tailored access to digital information (call centers, help systems, interactive banking and booking facilities, routing systems, information retrieval, etc.).</Paragraph>
    <Paragraph position="1"> The innovative goal of the Finnish Interact project is to enable natural language interaction in a wider range of situations than has been possible so far, and in situations where its use has not been functional or robust enough. This means that the systems should support rich interaction and also be able to learn and adapt their functionality to the changing situation. It also implies that the needs of special groups will be taken into account when designing more natural interactive systems. Within the current system, such scenarios can e.g. include an intelligent bus-stop which allows spoken and text interaction concerning city transportation, with a sign language help facility.</Paragraph>
    <Paragraph position="2"> The project addresses especially the problem of adaptivity: the users are situated in mobile environments in which their needs, activities and abilities vary. To allow the users to express their wishes in a way characteristic to them and Philadelphia, July 2002, pp. 64-73. Association for Computational Linguistics. Proceedings of the Third SIGdial Workshop on Discourse and Dialogue, the situation, interaction with the system should take place in a robust and efficient manner, enabling rich and flexible communication. Natural language is thus the preferred mode of interaction, compared to graphical interfaces for example. Adaptivity also appears in the techniques and methods used in the modelling of the interaction and the system's processing capabilities. An important aspect in this respect is to combine machine learning techniques with rule-based natural language processing, to investigate limitations and advantages of the two approaches for language technology.</Paragraph>
    <Paragraph position="3"> In this paper we focus on adaptivity which  The paper is organized as follows. We first introduce the dialogue system architecture. We then explain how the modules function and address the specific design decisions that contribute to the system's adaptivity. We conclude by discussing the system's capabilities and providing pointers for future work.</Paragraph>
  </Section>
class="xml-element"></Paper>