Lynn Lambert, \Recognizing Complex Discourse Acts: A Tripartite PlanBased Model of Dialogue," Technical Report 93-19, University of Delaware, Department of Computer and Information Sciences, Newark, Delaware, May 1993, PhD thesis. 49  Home/Search   Document Not in Database   Summary   Related Articles   Check

....categorizing the types and levels of plans that can exist in dialogue. We first discuss the work of Lambert, and then that of Ramshaw. We then look at further refinements by Carberry et al. Ardissono et al. and finally Chu Carroll and Carberry. Lambert Lambert ( Lambert and Carberry, 1991; Lambert, 1993] proposed a three level model of dialogue, consisting of the domain and meta levels of Litman and Allen as well as a level of discourse plans, which specifically handled recognition of multi utterance speech acts. The follow utterances, for example, constitute a warning (a discourse plan) only ....

Lynn Lambert, "Recognizing Complex Discourse Acts: A Tripartite PlanBased Model of Dialogue," Technical Report 93-19, University of Delaware, Department of Computer and Information Sciences, Newark, Delaware, May 1993, PhD thesis. 49

....one hypothesis from one component (the parser, for instance) into the next (the discourse processor) although in some cases, and to avoid an unmanageable number of hypothesis, local pruning is performed. The discourse module is based on Lambert s tripartite model [Lambert and Carberry 92] Lambert 93] It disambiguates the speech act of each sentence, normalizes temporal expressions and incorporates the sentence into a discourse plan tree, at the same time updating a calendar to keep track of the dates being discussed by the speakers. The final disambiguation combines a series of scores from ....

L. Lambert. Recognizing Complex Discourse Acts: A Tripartite PlanBased Model of Dialogue. PhD Thesis, University of Delaware, 1993.

.... of the knowledge which it is based on (Cohen Levesque, 1990; Cohen Perrault, 1979) while others had a more computational bias (Carberry, 1988) More recently domain plans have been complemented with higher levels plans called discourse plans (Litman Allen, 1987) and problem solving plans (Lambert, 1993). While Litman and Allen s discourse plans dealt both with communication strategies and problem solving activities, Lambert separates the discourse level in two parts: in her framework, communicating strategies are represented in the communicative level, while problem solving plans model the ....

Lambert, L. (1993). Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue.

....context list and becomes the current context. Initially, the list contains only the empty context, not specifying any values, thus matching every utterance. We are currently considering a more advanced context selecting mechanism, which is based on the focusing method of Ros e et al. 1995) and Lambert (1993). It is based on the use of a tree, in which the order of the branches defines the saliency of a context. 6 Concluding Remarks In this paper we have reported on some of the first steps we have taken towards the design of dialogue systems. First we presented an outline of the preprocessor of our ....

Lambert, L. (1993). Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue.

....partners about how the execution of an action proceeds. The main difference with respect to other planners is that we chose to describe the planner in a plan based declarative language. The plan based architecture allows us to apply the plan recognition algorithms developed so far (Carberry 1990b, Lambert 1993) in order to recognize and trace both the object level plans carried on by the agent, and 2 The term reactive planner is used here to denote reactive deliberative systems, also defined as practical reasoning systems (Jennings et al. 1998) on the contrary, we are not referring to the ....

.... plans, the speaker infers that the partner will also aim 8 Many subdialogues addressing the believability of statements can result from the failure of the Update action; for example, the execution of Update can detect the presence of a previous belief incompatible with the new information (Lambert 1993): A: CS144 is taught by Prof. Smith B: But wasn t Prof. Smith on sabbatical at other goals (not only concerning knowledge) because they are crucial to his plans. For instance, compare the two alternative responses in the following example: Example 4.2.1 A: Could you give me a pass for the CS ....

[Article contains additional citation context not shown here]

Lambert, L. (1993). Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. Ph.D. thesis, University of Delaware.

....level of meaning. 6 Related work First of all, some comments must be made about our representation of the context of a dialogue: our Context Models (indicated as CTXi in the figures) are rather different from the dialogue model described in Lambert Carberry (1991) Carberry et al. 1992) and Lambert (1993). As discussed in Ardissono et al. 1996) although both frameworks contain domain, communicative and metalevel plans, we make a different use of our (AM) metalevel and communicative plans. In Lambert s model, the problem solving plans relate communicative actions to the domain level plans pursued ....

L. Lambert (1993). Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. Ph.D. thesis, University of Delaware.

....can be associated with more than one inference chain, representing different possible speech acts which could be performed by the utterance out of context. Focusing heuristics are used to rank the different inference chains and choose the one which attaches most coherently to the discourse context [3, 8]. However, since heuristics can make wrong predictions, the speech act may be misrecognized, thus making the context inaccurate for future context based predictions. Unexpected input, disfluencies, out of domain utterances, and missing information add to the frequency of misrecognition in ....

....act recognition in Section 4. Lastly, we evaluate the effects of the proposed approaches on reducing cumulative error. 2 Related Work There has been much recent work on building a representation of the discourse context with a plan based or finite state automaton based discourse processor [1, 9, 3, 7, 8, 5]. Of these, the Verbmobil discourse processor [7] and our Enthusiast discourse processor are designed to be used in a wide coverage, large scale, spontaneous speech system. In these systems, the design of the dialogue model, whether plan based or a finite state machine, is grounded in a corpus ....

L. Lambert. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. PhD thesis, Department of Computer Science, University of Delaware, 1993.

....discourse plans) which have other plans as parameters. Using these notions, Litman and Allen are able to account for a larger range of utterances than previous plan based approaches, including sub dialogues to clarify or correct deficiencies in a plan under discussion. Carberry and Lambert [25, 64, 63], make a further distinction, adding problem solving plans to the discourse and domain plans of Litman and Allen. Problem solving plans are ways of building up actual domain plans, and are usually achieved by way of discourse plans. Cohen and Levesque extend the work described in Section 3.3.1, ....

Lynn Lambert. Recognizing Complex Discourse Acts: A Tripartite PlanBased Model of Dialogue. PhD thesis, University of Delaware, 1993. Reproduced as TR 93-19 Department of Computer and Information Science, University of Delaware.

....knowledge intensive processes, such as recognizing doubt or clarification sub dialogues and robust ellipsis resolution. 3 Plan inference based discourse processing Our plan based discourse processor follows previous work of using a plan inference model for understanding discourse utterances [2, 27, 7, 22]. Plan based approaches are based on Austin s [3] and Searl s [40] notionof language as a series of actions, and Grice s [13] and Searle s description of language as goal oriented behavior. Specifically, we developed our discourse processor following Lambert s work [22, 23] on understanding ....

....utterances [2, 27, 7, 22] Plan based approaches are based on Austin s [3] and Searl s [40] notionof language as a series of actions, and Grice s [13] and Searle s description of language as goal oriented behavior. Specifically, we developed our discourse processor following Lambert s work [22, 23] on understanding negotiation dialogues. Lambert s work extends previous work on discourse processing in several ways: 1) separating three levels of processing: the domain level, the problem solving level and the discourse level; 2) recognizing actions expressed over multiple sentences; and (3) ....

[Article contains additional citation context not shown here]

Lynn Lambert. Recognizing complex discourse acts: A tripartite plan-based model of dialogue. Technical Report Technical Report No. 93-19, Department of Computer and Information Sciences, University of Delaware, 1993.

....dominio. Nel lavoro di Litman e Allen [11] sono stati usati piani di discorso per modellare sia gli aspetti linguistici del dialogo, che quelli associati al comportamento di problem solving. Lambert ha separato la doppia funzione dei piani di discorso di Litman e Allen in due attivita distinte [10]. Tuttavia, i suoi piani di problem solving non trattano in modo uniforme i piani di dominio e quelli linguistici: Lambert usa il livello di problem solving per specificare le relazioni tra le frasi in input e le azioni di dominio, ma non per modellare il comportamento linguistico. Quest ultimo e ....

....un modello generale di interpretazione del dialogo. Nell articolo si e mostrato che una rappresentazione esplicita e dichiarativa del processo di problemsolving e un elemento fondamentale nello sviluppo di un sistema di dialogo. Anche se la nozione di piano di problem solving non e nuova ([15, 10]) noi usiamo questi piani in una prospettiva nuova, per costruire un agente coinvolto nel dialogo in modo diretto, con i propri piani e obiettivi. In questa prospettiva, i piani di PS sono usati per modellare l interazione tra agenti a livello linguistico e di dominio. Ciascun agente si comporta ....

L. Lambert. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. PhD thesis, University of Delaware, 1993.

.... background information is used to justify the subsequent speech acts, or to check whether the preconditions of the actions which should be performed next are met (pre sequences) Many plan recognition systems interpret input sentences without performing a discourse analysis ( 5, 1] others (e.g. [8]) perform a basic one, but they still don t pay attention enough to the communicative role of sentences; in general, all of them use speech acts for identifying the speaker s domain plans, without considering the fact that, sometimes, utterances are too general to address particular actions in the ....

L. Lambert. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. PhD thesis, University of Delaware, 1993.

....the user s domain goals, which the user will later execute. Research on plan recognition within the expert consultation dialogue framework has focused on inferring the user s domain goals and discourse actions, for example [ Allen and Perrault, 1980; Litman and Allen, 1987; Carberry, 1987; Lambert, 1993 ] However, in order to perform effectively as an expert consulting agent, such a system s plan recognition ability must be augmented with several other types of knowledge. Specifically, the system must be able to 1) recognize the user s preferences regarding the domain plan being constructed and ....

Lynn Lambert. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. PhD thesis, University of Delaware, Newark, Delaware, 1993.

....Processing The discourse processing module in Janus disambiguates the speech act of each SDU, updates a dynamic memory of schedules, and incorporates the SDU into discourse context. We have experimented with two approaches to discourse processing: a plan inference system (based on work by Lambert [1]) and a finite state processor Unsegmented Speech Recognition: noise si1 mira toda la man5ana estoy disponible noise noise y tambie1n el fin de semana si podri1a hacer mejor un di1a fin de semana porque justo el once no puedo me es imposible va a poder fin de semana) Pre broken Speech ....

L. Lambert. RecognizingComplex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. PhD thesis, Department of Computer Science, University of Delaware, 1993.

....when an incorrect hypothesis is chosen and incorporated into the context, thus providing an inaccurate context from which subsequent context based predictions are made. For example, in Enthusiast, we model the discourse context using speech acts to represent the functions of dialogue utterances [1, 3, 6, 7]. Speech act selection is strongly related to the task of determining how the current input utterance relates to the discourse context. When, for instance, a plan based discourse processor is used to recognize speech acts, the discourse processor computes a chain of inferences for the current ....

....can be associated with more than one inference chain, representing different possible speech acts which could be performed by the utterance out of context. Focusing heuristics are used to rank the different inference chains and choose the one which attaches most coherently to the discourse context [3, 8, 5]. However, since heuristics can make wrong predictions, the speech act may be misrecognized, thus making the context inaccurate for future context based predictions. Unexpected input, disfluencies, out of domain utterances, and missing information add to the frequency of misrecognition in ....

[Article contains additional citation context not shown here]

L. Lambert. Recognizing Complex Discourse Acts: A Tripartite PlanBased Model of Dialogue. PhD thesis, Department of Computer Science, University of Delaware, 1993.

....approach with those from the non context based parser approach. In developing our discourse processor for disambiguation we needed to address three major issues. First, most plan based or finite state automaton based discourse processors (Allen and Schubert, 1991; Smith, Hipp, and Biermann, 1995; Lambert, 1993; Reithinger and Maier, 1995) including the one we initially developed (Ros e et al. 1995) only take one semantic representation as input at a time: thus, we had to extend the discourse processor so that it can handle multiple hypotheses as input. Secondly, we needed to quantify the ....

.... suggest accept) who ( frame i) frame free) sentence type state) Sentence: I could do it Wednesday morning too. Figure 1: An Example ILT The resulting set of ILTs is then sent to the discourse processor. The discourse processor, based on Lambert s work (Lambert and Carberry, 1992; Lambert, 1993), disambiguates the speech act of each sentence, normalizes temporal expressions from context, and incorporates the sentence into the discourse context represented by a plan tree. The discourse processor also updates a calendar which keeps track of what the speakers have said about their ....

[Article contains additional citation context not shown here]

Lambert, L. 1993. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. Ph.D. thesis, Department of Computer Science, University of Delaware.

....ILTs from each sub lattice are combined, yielding a list of ILT sequences that represent the possible sentences of a full multi sentence turn. An ILT n gram is applied to each such list to determine the probability of each sequence of sentences. The discourse processor, based on Lambert s work ([14, 15]) disambiguates the speech act of each sentence, normalizes temporal expressions, and incorporates the sentence into a discourse plan tree. The discourse processor s focusing heuristics and plan operators eliminate some ambiguity by filtering out hypotheses that do not fit into the current ....

....was evaluated. As can be seen in Table 2, the combined disambiguation method succeeds in selecting the correct parse in 209 of the 252 cases, a success rate of 82 . Disambiguation Using Discourse Plans Our discourse processor is a plan inference model based on the recent work of Lambert ([14, 15]) The system takes as its input ILTs of sentences as they are uttered and relates them to the existing context, i.e. the plan tree. Plan inferencing starts from the surface forms of sentences. Then speech acts are inferred. Multiple speech acts for one ILT could be inferred. A separate inference ....

L. Lambert. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Models of Dialogue. PhD thesis, University of Delaware, September 1993.

....texts, or ILTs, representing the semantic content and structure of the clauses of the utterance that were parsed by the parser. The GLR parser can output either a single best disambiguated parse result, or a ranked list of possible analyses. The discourse processor [81] based on Lambert s work [51] [52] disambiguates the speech act of each sentence, normalizes temporal expressions, and incorporates the sentence into a discourse plan tree. The discourse processor s focusing heuristics and plan operators eliminate some ambiguity by filtering out hypotheses that do not fit into the current ....

L. Lambert. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. PhD thesis, University of Delaware, Newark, DE, September 1993.

....beliefs. This Combine Belief action provides an abstract model for an action taken by the hearer rather than by the speaker. A complete model of the manner in which a user combines belief from several utterances is beyond the scope of this paper. Further formal work in this area, such as that in [10] is essential for accurately representing the structure of discourse plans. DPOCL provides a 2 Subplans in this figure are grouped inside rounded boxes for ease of reference. 3 DPOCL uses the standard POCL technique of encoding the initial conditions and the goals of a planning problem as the ....

Lynn Lambert. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. PhD thesis, University of Delaware, 1993.

No context found.

Lynn Lambert, \Recognizing Complex Discourse Acts: A Tripartite PlanBased Model of Dialogue," Technical Report 93-19, University of Delaware, Department of Computer and Information Sciences, Newark, Delaware, May 1993, PhD thesis. 49

No context found.

Lambert, L. (1993). Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. Doctoral dissertation, Department of Computer and Information Sciences, University of Delaware, Technical Report 93-19..

No context found.

Lambert, L. 1993. "Recognizing complex discourse acts: a tripartite plan-based model of dialogue." Ph.D. dissertation. Technical Report 93-19, Department of Computer and Information Sciences, Univerity of Delaware.

No context found.

Lambert, Lynn. 1993. Recognizing Complex Discourse Acts: A Tripartite Plan-Based Model of Dialogue. PhD. Dissertation. The University of Delaware.

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