<?xml version="1.0" standalone="yes"?>
<Paper uid="C92-1053">
  <Title>ORGANIZING DIALOGUE FROM AN INCOHERENT STREAM OF GOALS*</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
I INTRODUCTION
</SectionTitle>
    <Paragraph position="0"> Conversation seems coherent and is easy to follow because it reflects the way people think. When the speaker thinks coherently, his or her communication goals will be properly organised to follow linguistic convention. So s models of human language generation can allow domain goals to directly motivate conversation and add clue words only when the occasional utterance violates convention \[5; 8\].</Paragraph>
    <Paragraph position="1"> However~ computer-generated conversation cannot rely on problem solving for its organisation. Some problem solvers make no attempt to be &amp;quot;cognitively plausible&amp;quot; and do not produce goals in sequences that would appear coherent to human users. The combined goals from a distributed problem solver where several independent reasoners use a single interface to communicate with the user are also likely to be incoherent. Even if individual problem solvers produce coherent streams of goals, the stream of goals from the aggregate is likely to switch back and forth between sub-problems that are being addressed by the individual systems. We call * Some of the work described in this paper was done at the Computer Science Depaxtment of Georgia Institute of Technology, AtlJnts, Georgia, 30332, supported by the NSF under llrmat lST-S608362. The &amp;uther wishes to thank the anonymous reviewers for their comments.</Paragraph>
    <Paragraph position="2"> the sequence of goals produced by such systems an incoherent stream of goab because the goals are ordered in a way that would not seem reasonable to a human listener. Interfaces to such systems, while being responsive to the goals of problem solving, must rely on something else to give dialogue its organisation.</Paragraph>
    <Paragraph position="3"> In this paper, we describe a template that can provide computer-generated conversation with a coherent organisation while meeting the needs of the problem solvers. This template is built from schemata representing expected discourse structure. These schemata include general linguistic conventions as well as expected organisations for specific situations. Before conversation begins the template is very abstract, giving only broad descriptions of topics that might be expected in a specific domain. These expectations provide a framework for organising goals from the problem solvers. New goals are associated with the existing template, organising them into groups that are related by conventional discourse structure. Goals are added by finding schemata which connect them to the template. Since the schemata specify acceptable conversational structure, the template represents a coherent conversation. The dialogue in Figure 1 was organised using such a template. The arrows in the figure show the order in which goals arrived at the interface. The tall of the arrow shows when a goal arrived at the interface; the head, where it is realised in the dialogue. Each time the arrows cross, the goal has been delayed to fit more naturally into the conversation. Goals have been grouped by the course that they relate to as well as being further organised by attributes such as ingredients of the main-dish of a course (utterances 10a~10d). The dialogue also includes a story to connect two utterances (10a and 10b), which adds variety to the structure of the conversation.</Paragraph>
    <Paragraph position="4"> This dialogue was generated by JUDIS \[15\], an interface to Julia, an advisor for meal-planning \[3\]. I Jut JUDIS is rcsponsihle for the organlsing the conversation mad works with conceptl instead of actual English ut ter~c~s. Concepts are converted to Ensllsh and En~ah trandated into the conceptual language by the NLPTool \[2\]. JUDIS ~ worked with the NLPTool to participate in Engiish dialogues, hut wu not col~ected to the NLPTool for this C/x~ple. In the fillur~, EnsUsb is used to make the dialogue readable.</Paragraph>
    <Paragraph position="5"> The Julla described here is the original version of the system. AcrEs DE COLING-92, NANTES, 23-28 AOt3T 1992 3 3 8 PRec. OF COLING-92, NANTES, AUO. 23-28. 1992  (1) User: i'la planning a dinner and would like to ~erve Mexclan food.</Paragraph>
    <Paragraph position="6"> (2) JUDIS:I'm thinking of chili for a ll~ill course. D~s everyone llke spicy food? (3) User: No, Tom hates splcy food.</Paragraph>
    <Paragraph position="7"> ~(4) JUDIS:Would you like enchiladas? (5) Uses: What are enchiladas? ~(6) dODIS:They'ge a Mexlc~ food made of tortillas and rmsat. They also contain onion and garlic.</Paragraph>
    <Paragraph position="8"> (7) User: OK.</Paragraph>
    <Paragraph position="9"> (8) dOD~S:het'~ have guacarnole for at~ appetizer. ~ 191 User: OK.</Paragraph>
    <Paragraph position="10"> (IOakTUDIS: I ~ugt r~bered a case where q t sonmol3e had a party ~d could ,tot mash h tile avocados because they wore ultrib~. (lOb} Later. go--one could not eat the enchiladas because he wa~ allergic to I cileese.</Paragraph>
    <Paragraph position="11"> (10c) Make sure no one is allergic to cheese and that the av~ados at'e rlpe enough.</Paragraph>
    <Paragraph position="12"> (10d) You n~ay w~it to use red onions because they are sweet.</Paragraph>
    <Paragraph position="13"> (10o) Shall we g,z~e the appetizer at the table or as her d'oeuveres? (Ii) User: At the table. ~ (12) JUDIS: Would sopapllla or fried ice cre~ for dessert.? (13) User: SopapJlla.</Paragraph>
    <Paragraph position="14"> (14) JUDIS:  |just remembered a meal where soC/msone couldn't eat suaca~le because he was allergic to tcwnat~s. Are ally of your guests allergic to somalis? (15) User : No.</Paragraph>
    <Paragraph position="15">  lia is a distributed problem solver comprised of a case-based reasoner and a planner which uses more traditional problem solving techniques. The problem solvers and the interface share their world knowledge. When a problem solver has a goal to either give or get information, the goal is sent to JUDIS where it is converted into an appropriate utterance and attached to the template. We refer to these goal-motivated utterances as requests.</Paragraph>
  </Section>
class="xml-element"></Paper>