Susan McRoy. Abductive Interpretation and Reinterpretation of Natural Language Utterances. PhD thesis, University of Toronto, 1993. Reproduced as TR CSRI-288 Department of Computer Science, University of Toronto.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... challenges the utility of the deep intention recognition in [8] claiming instead that considering the possible actions and their immediate effects (including expectations set up for future utterances) is sufficient, especially when combined with a facility to repair erroneous conclusions [74, 75]. While recognizing the intentions and plans of a speaker can play a useful role in act interpretation, and may in fact be crucial to an agent s actual success in some domains, one must also remember the on line nature of dialogue in this case as well. More prominence should be placed on ....

Susan McRoy. Abductive Interpretation and Reinterpretation of Natural Language Utterances. PhD thesis, University of Toronto, 1993. Reproduced as TR CSRI-288 Department of Computer Science, University of Toronto.

....moreover that there may be a core set of meta dialog principles that is in some sense complete. If we are right, then implementing such a set would be of considerable interest. We give examples of existing computer programs that converse inadequately according to our guidelines. Introduction In (McRoy 1993) McRoy urges that the ability to deal with mistakes is central to communication, and moreover that it is best treated as part and parcel of the same reasoning and linguistic abilities as the rest of the communication process, rather than as a separate and optional refinement. We agree. In fact, ....

....of the paper. T1 Mother: Do you know who s going to that meeting T2 Russ: Who T3 Mother: I don t know. T4 Russ: Oh. Probably Mrs. McOwen and probably Mrs. Cadry and some of the teachers. Figure 1: Russ notices an inconsistency at T3 and makes a repair at T4 McRoy and Hirst McRoy and Hirst (McRoy 1993; Hirst McRoy 1995) consider misunderstanding and repairs in dialog. In their model of conversation, misunderstanding is signalled by an inconsistency between the expectations of a dialog participant and an utterance. The agent must then reason about and explain this inconsistency. This can lead ....

McRoy, S. 1993. Abductive Interpretation and Reinterpretation of Natural Language Utterances. Ph.D.

....2.7 Conversation Analysis Conversation Analysis is a subfield of sociology that is concerned with the study of actual dialogues in order to discover recurring patterns. 3 Work done in this field is starting to enter into computation approaches to natural language (e.g. Traum and Allen, 1994; McRoy, 1993), for it allows expectations and social norms to be used in understanding a speaker s utterance, thus short circuiting means end reasoning about the other participant s top level goals and beliefs, and in deciding what to do next. The contributions made in the field of conversation analysis are ....

McRoy, Susan W. 1993. Abductive interpretation and reinterpretation of natural language utterances. Doctoral dissertation, Technical Report CSRI 288, Department of Computer Science, University of Toronto.

....environment. The topic of repair in natural language conversation has been a topic of much study in Conversation Analysis and Psycholinguistics (e.g. Schegloff et al. 1977; Levelt, 1983; Clark and Schaefer, 1989] and more recently in AI (e.g. Litman and Allen, 1990; Raudaskoski, 1990; McRoy, 1993; Traum and Hinkelman, 1992] There has also been work in plan execution monitoring and replanning in an uncertain world (e.g. Peng Si Ow, 1988; Ambros Ingerson and Steel, 1988] but there has as yet been little foundational work on explicating formally just what errors and repairs are, and ....

Susan McRoy, Abductive Interpretation and Reinterpretation of Natural Language Utterances, PhD thesis, University of Toronto, 1993, Reproduced as TR CSRI-288 Department of Computer Science, University of Toronto.

....party will bring about an obligation to perform the requested action. If the obligation is to say something then we call this a discourse obligation. Our model of obligation is very simple. We use a set of rules that encode discourse conventions. Whenever a new conversation act is determined 1 [McRoy, 1993] uses expectations derived from Adjacency Pair structure [Schegloff and Sacks, 1973] as are many of the discourse obligations consideredin this paper. Theseexpectationscorrespond to social norms and do impose the same notion of accountability. However, the analysis there is oriented towards ....

Susan McRoy, Abductive Interpretation and Reinterpretation of Natural Language Utterances, PhD thesis, University of Toronto, 1993, Reproduced as TR CSRI-288 Department of Computer Science, University of Toronto.

.... Allen, 1987) while other models consider the plans and goals at a level lower than the surface utterance plans (Carletta, 1990a; Carletta, 1990b; Grosz and Sidner, 1990; Lochbaum, Grosz and Sidner, 1990; Lambert and Carberry, 1991) A few models have considered both generation and understanding (McRoy, 1993; Heeman and Hirst, 1992) Heeman and Hirst s system models both plan construction and inference while accounting for how speech acts correspond to an utterance. However, their system, because it is for generating and understanding referring expressions (with only one underlying intention) cannot ....

McRoy, S. W. (1993). Abductive Interpretation and Reinterpretation of Natural Language Utterances. PhD thesis, University of Toronto.

.... psychological constraints and the limited resources that humans have, it is conceivable that incremental processing is impossible without backtracking this would be consistent with the mistakes and re interpretations that are observed in naturally occurring conversations [ Hirst et al. 1993, McRoy, 1993 ] 5 Studying the ways in which the algorithms presented in this chapter can be modified in order to derive valid text structures incrementally is, however, outside the scope of this thesis. 5 I thank Graeme Hirst for bringing up this hypothesis. 3.7 Summary In this chapter, I have ....

Susan W. McRoy. Abductive Interpretation and Reinterpretation of Natural Language Utterances. PhD thesis, Department of Computer Science, University of Toronto, 1993.

....if they are poorly written or expressed; highquality language generation is very difficult for people. Nevertheless, people are, in general, quite successful in their use of language. That s because they have strategies This paper summarizes work reported in greater detail in references [15, 20, 21, 22, 27, 28]. y Susan McRoy is now at the Department of Electrical Engineeringand Computer Science, University of Wisconsin Milwaukee, Milwaukee, WI 53201, U.S.A. z Peter Heeman is now at the Department of Computer Science, University of Rochester, Rochester, NY 14627, U.S.A. for coping with their ....

....of particular participants. What these accounts lack that computational accounts provide is an explanation of how people can identify the convention that is relevant, especially when there is no pre existing expectation. 5. 3 A synthesis In our work (described more fully in references [27, 28]) we have developed a model of interaction that addresses the possibility that the participants might differ about the speech act that is performed by some utterance, without requiring extended reasoning about the speaker s goals. According to the model, speakers form expectations on the basis of ....

McRoy, Susan W. Abductive interpretation and reinterpretation of natural language utterances. PhD thesis, published as technical report CSRI-288, Department of Computer Science, University of Toronto, 1993.

....as a coherent attempt to initiate a conversational exchange if: ffl There is some action that the system wants the user to do. 9 The definitions would also apply from the user s point of view; however, her interpretations would differ from the system s when there had been a misunderstanding. McRoy [ 1993 ] discusses this in detail, showing the representations that result when two copies of the system communicate with each other. ffl This action would be the expected response to the utterance. There is an additional coherence constraint that both the beliefs that are expressed by the system s ....

Susan W. McRoy. Abductive Interpretation and Reinterpretation of Natural Language Utterances. PhD thesis, Department of Computer Science, University of Toronto, Toronto, Canada, 1993. Available as CSRI Technical Report No. 288, University of Toronto, Department of Computer Science.

No context found.

Susan W. McRoy (1993), Abductive interpretation and reinterpretation of natural language utterances, PhD thesis, published as technical report CSRI-288, Department of Computer Science, University of Toronto.

....need to have some idea of the beliefs and intentions that particular actions can express, so that they can make judgments about their appropriateness in the context. Thus, some synthesis of Conversation Analysis and intentional approaches appears to be necessary. A possible synthesis In our work [16, 17] we have developed a model of communicative interaction that supports the negotiation of meaning. According to the model, speakers form expectations on the basis of what they hear, and thus monitor for differences in understanding. If necessary, they also reinterpret utterances in response to new ....

Susan W. McRoy. Abductive Interpretation and Reinterpretation of Natural Language Utterances. PhD thesis, Department of Computer Science, University of Toronto, Toronto, Canada, 1993. Published as CSRI Technical Report No. 288, University of Toronto, Department of Computer Science.

.... In this paper, we shall describe an abductive account of interpreting speech acts and recognizing misunderstandings (we discuss the generation of repairs of misunderstandings in McRoy and Hirst, 1992) This account is part of a unified theory of speech act production, interpretation, and repair [ McRoy, 1993 ] According to the theory, speakers use their beliefs about the discourse context and which speech acts are expected to follow from a given speech act in order to select one that accomplishes their goals and then to produce an utterance that performs the chosen speech act. Interpretation and ....

Susan W. McRoy. Abductive Interpretation and Reinterpretation of Natural Language Utterances. PhD thesis, Department of Computer Science, University of Toronto, Toronto, Canada, 1993. In preparation.

No context found.

McRoy, Susan W. 1993a. Abductive Interpretation and Reinterpretation of Natural Language Utterances. Ph.D. thesis, Department of Computer Science, University of Toronto, Toronto, Canada. Published as CSRI Technical Report No. 288, University of Toronto, Department of Computer Science.

....can be addressed without the unconstrained inference of goals. This is possible because in situations where communication has broken down, we can repair the trouble by identifying how it has broken down, without performing extended reasoning as to why. 2 The structured intentional approach [ McRoy, 1993; McRoy, 1995; McRoy and Hirst, To appear ] presents a model of communicative interaction that addresses the possibility that the participants might differ about the speech act that is performed by some utterance, without requiring extended reasoning about the speaker s goals. This work formalizes ....

....It may pursue any strategy so long as its conditions are true (or assumable by default) and the intentions conventionally expressed by an act are consistent with any previously expressed attitudes. It then selects an utterance that is known to be a possible realization for the act. Although [ McRoy, 1993 ] does not consider the problem of indirect utterances, the model is being extended to address them. This extension will include abductive plan based inference as a subtask, guided by the hearer s own goals and expectations. Such an extension will provide an interface between discourse reasoning ....

Susan W. McRoy. Abductive Interpretation and Reinterpretation of Natural Language Utterances. PhD thesis, Department of Computer Science, University of Toronto, Toronto, Canada, 1993. Published as CSRI Technical Report No. 288, University of Toronto, Department of Computer Science.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC