<?xml version="1.0" standalone="yes"?>
<Paper uid="W97-0611">
  <Title>I I DISTANCE ZONE I \ %alce=-=ource takeo-de=tinatPSon I I SOURCE DESTINATION I TOTAL-COST / I / I</Title>
  <Section position="4" start_page="55" end_page="55" type="metho">
    <SectionTitle>
4 Planning an utterance
</SectionTitle>
    <Paragraph position="0"> So far, we have described possible dialogue continuations and interpreted them in the context of the dialogue history as pursuing a particular communicative goal. In most current dialogue systems, each of these goals would be realized as a separate utterance, i.e. the surface structure of the dialogue would merely be a reflection of the underlying dialogue history (see dialogue 1). Our goal, however, is to generate utterances like those in sample dialogue 2. Hence, we need to investigate which communicative goals can be satisfied at a time, in other words, which constellations of dialogue acts given a certain state of the task model can be jointly expressed in one utterance. Recent work on aggregation in the context of natural language generation (e.g.(Dalianis and Hovy 1993)) states that surface structures are abbreviated when information units that in the domain are represented as separate individuals share pertinent features, for instance, syntactical, lexical or semantic features. We extended this notion to allow aggregation of communicative goals: Depending on the common feature, we defined four strategies for condensing dialogue interaction: abbreviation, abstraction, omission, and dominance. null Abbreviation. We call the condensing of information &amp;quot;abbreviation&amp;quot; when a number of continuations that would become adjoining parts of the parse tree and furthermore represent the same communicative goal are expressed in one utterance. If the interhal structure of adjoining dialogue cycles (siblings, see Figure 4a) is identical, and the concepts/tasks negotiated in these structures either have the same superconcept or are connected to the same concept by means of a relation, and if the state of the concepts under consideration is open, then the resulting utterance is abbreviated. For instance, abbreviation of several offer constellations that represent the initiate choice as in &amp;quot;Do you want the rate or the total cost of a call?&amp;quot;, where 'acquire uval(TESADIS) * sub(TESADIS)' = 'acquire uval 0 * inst(PRICE-PER-MINUTE) t..J inst(TOTAL-COST)' are abbreviated. Another example would be the abbreviation of acts for acquiring specific information as in &amp;quot;What are the source and the destination of your call?&amp;quot;, where 'acquire uval(source)' and 'acquire uval(destination)' are aggregated. Figure 4b, which could be a continuation of the Accept(u) in Figure 4a, illustrates this case.</Paragraph>
    <Paragraph position="1"> A third example is abbreviation of several acquire confirmation goals, e.g., &amp;quot;Do you want to call Darmstadt from Magdeburg?&amp;quot;--an abbreviation of 'acquire con- null mation sval(destination) = uval(destination)'.</Paragraph>
    <Paragraph position="2"> Abstraction means that we transform several similar goals into a new, more abstract goal. In particular, this applies to the goal 'acquire disambiguation' when a large number of alternatives are at hand. Research in cognitive science and ergonomic design of dialogue systems have shown that human beings can only keep a few alternatives in their short term memory, hence instead of presenting the listener with a long list of alternatives, it is more efficient to phrase a question in a way that avoids mentioning the alternatives. For instance, the goals 'acquire disambiguation sval(name) = maier', 'acquire disambiguation svai(name) = meyer', etc. can be abstracted into 'acquire disambiguation sval(name) in maier,meyer,meier,mayer' with the realization &amp;quot;How do you spell \[mai:er\]?&amp;quot; instead of &amp;quot;Is \[mai:er\] spelt with a i, a y, e i, or e y?&amp;quot; Dominance and subordination. It can be efficient to solicit implicit confirmation of previously recognized values, hence we allow a goal for acquiring a new value and a goal for acquiring confirmation of another value to be realized in one utterance, as in &amp;quot;When do you want to call Frankfurt?&amp;quot; This is a realization of the dominating goal 'acquire uval(startTime)' and the subordinate goal 'acquire confirmation of svai(destination)'.</Paragraph>
    <Paragraph position="3"> Omission. Omission means that we leave out a goal altogether, for instance, if the recognition rate of an ambiguous value is high, we take the risk of asking for disambignation right away, as in the question &amp;quot;Is your call from Frankfurt am Main or Frankfurt an der Oder?&amp;quot; Here the goal 'acquire confirmation sval(source) = uval(source)' is omitted and the goals 'acquire disambiguation of sval(destination) = Frankfurt am Main' 'acquire disambiguation of sval(destination)= Frankfurt an der Oder' have been abstracted as above.</Paragraph>
    <Paragraph position="4"> When is aggregation not possible? Above we discussed which structures can be successfully aggregated into more abstract goals and more compact utterances. However, there are certain limits to performing aggregation. For instance, one cannot aggregate between different levels in the dialogue history if the higher level has not yet been satisfied as the following two examples illustrate: *&amp;quot;Do you want the rate or the total cost of a call to where? or ?&amp;quot;When do you want the rate of a call?&amp;quot;.</Paragraph>
  </Section>
class="xml-element"></Paper>