<?xml version="1.0" standalone="yes"?>
<Paper uid="P90-1010">
  <Title>Mixed Initiative in Dialogue: An Investigation into Discourse Segmentation</Title>
  <Section position="2" start_page="0" end_page="70" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Conversation between two people has a number of characteristics that have yet to be modeled adequately in human-computer dialogue. Conversation is BIDIRECTIONAL; there is a two way flow of information between participants. Information *This research was partially funded by ARO grants DAAG29-84-K-0061 and DAAL03-89-C0031PRI, DARPA grant N00014-85-K0018, and NSF grant MCS-82-19196 at the University of Pennsylvania, and by Hewlett Packard, U.K.</Paragraph>
    <Paragraph position="1"> is exchanged by MIXED-INITIATIVE. Each participant will, on occasion, take the conversational lead. Conversational partners not only respond to what others say, but feel free to volunteer information that is not requested and sometimes ask questions of their own\[Nic76\]. As INITIATIVE passes back and forth between the discourse participants, we say that CONTROL over the conversation gets transferred from one discourse participant to another. Why should we, as computational linguists, be interested in factors that contribute to the interactivity of a discourse? There are both theoretical and practical motivations. First, we wish to extend formal accounts of single utterances produced by single speakers to explain multi-participant, multi-utterance discourses\[Po186, CP86\]. Previous studies of the discourse structure of multi-participant dialogues have often factored out the role of MIXED-INITIATIVE, by allocating control to one participant\[Gro77, Coh84\], or by assuming a passive listener\[McK85, Coh87\]. Since conversation is a collaborative process\[CWG86, SSJ74\], models of conversation can provide the basis for extending planning theories\[GS90, CLNO90\]. When the situation requires the negotiation of a collaborative plan, these theories must account for the interacting beliefs and intentions of multiple participants. ~,From a practical perspective, there is ample evidence that limited mixed-initiative has contributed to lack of system usability. Many researchers have noted that the absence of mixed-initiative gives rise to two problems with expert systems: They don't allow users to participate in the reasoning process, or to ask the questions they want answered\[PHW82, Kid85, FL89\]. In addition, question answering systems often fail to take account of the system's role as a conversational partner.</Paragraph>
    <Paragraph position="2">  For example, fragmentary utterances may be interpreted with respect to the previous user input, but what users say is often in reaction to the system's previous response\[CP82, Sid83\].</Paragraph>
    <Paragraph position="3"> In this paper we focus on interactive discourse.</Paragraph>
    <Paragraph position="4"> We model mixed-initiative using an utterance type classification and a set of rules for transfer of control between discourse participants that were proposed by Whittaker and Stenton\[WS88\]. We evaluate the generality of this analysis by applying the control rules to 4 sets of dialogues, including both advisory dialogues (ADs) and task-oriented dialogues (TODs). We analysed both financial and support ADs. The financial ADs are from the radio talk show &amp;quot;Harry Gross: Speaking of Your Money &amp;quot;1 The support ADs resulted from a client phoning an expert to help them diagnose and repair various software faults ~. The TODs are about the construction of a plastic water pump in both telephone and keyboard modality S.</Paragraph>
    <Paragraph position="5"> The application of the control rules to these dialogues lets us derive domain-independent discourse segments with each segment being controlled by one or other discourse participant. We propose that control segments correspond to different subgoals in the evolving discourse plan. In addition, we argue that various linguistic devices are necessary for conversational participants to coordinate their contributions to the dialogue and agree on their mutual beliefs with respect to a evolving plan, for example, to agree that a particular subgoal has been achieved. A final phenomenon concerns shifts of control and the devices used to achieve this. Control shifts occur because it is unusual for a single participant to be responsible for coordinating the achievement of the whole discourse plan. When a different participant assumes control of a discourse subgoal then a control shift occurs and the participants must have mechanisms for achieving this.</Paragraph>
    <Paragraph position="6"> The control framework distinguishes instances in which a control shift is negotiated by the participants and instances where one participant seizes control.</Paragraph>
    <Paragraph position="7"> This paper has two objectives: 110 randomly selected dialogues (474 turns) from a corpus that was collected and transcribed by Martha Pollack and Julia Hirschberg\[HL87, PHW82\].</Paragraph>
    <Paragraph position="8"> 24 dialogues (450 turns) from tapes made at one of Hewlett-Packard's customer response centers. See \[WS88\]. 35 keyboard (224 turns) and 5 telephone dialogues (714 turns), which were collected in an experiment by Phil Cohen to explore the relationship between modality, interactivity and use of referring expressions\[Coh84\].</Paragraph>
    <Paragraph position="9"> To explore the phenomenon of control in relation to ATTENTIONAL STATE \[GS86, GJW86, Sid79\] 4. We predict shifts of attentional state when shifts in control are negotiated and agreed by all participants, but not when control is seized by one participant without the acceptance of the others. This should be reflected in different distribution of anaphora in the two cases.</Paragraph>
    <Paragraph position="10"> To test predictions about the distribution of control in different types of dialogues. Because the TOD's embody the master-slave assumption\[GSg0\], and control is allocated to the expert, our expectation is that control should be located exclusively with one participant in the TODs in contrast with the ADs.</Paragraph>
  </Section>
class="xml-element"></Paper>