Lars Ahrenberg, Arne J onsson, and Nils Dahlb ack. Discourse representation and discourse management for a natural language dialogue system. Technical report, Institutionen f or Datavetenskap, Universitetet och Tekniska H ogskolan Link oping, 1991.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... Local Dialogue Structure There have been several proposals for the kind of dialogue unit represented in (1) using structural terms like adjacency pair [ Scheglo and Sacks, 1973 ] exchange [ Sinclair and Coulthard, 1975 ] game [ Severinson Eklundh, 1983, Carletta et al. 1997 ] IR unit [ Ahrenberg et al. 1990 ] and argumentation act [ Traum and Hinkelman, 1992 ] At an abstract level, we need an IRF unit which can contain the three moves or acts: Initiative, Response, and Feedback, as indicated in (2) 2) 1] Initiative: Request(Act) Instruct] 2] Response: Do(Act) 3] Feedback: Eval [ ....

Lars Ahrenberg, Nils Dahlback, and Arne Jonsson. Discourse representation and discourse management for a natural language dialogue system. In Proceedings of the Second Nordic Conference on Text Comprehension in Man and Machine, 1990. 27

....along with social correlates such as mutual belief, joint intention, and obligation) e.g. Bretier Sadek, 1996; Traum Allen, 1994) or one can take a more structural view of the dialogue, concentrating on the performance of actions and various sorts of accessibility relationships. e.g. (Ahrenberg et al. 1990)) It may also be useful to distinguish components of information state into static and dynamic aspects. The former are those aspects of information state that are not expected to change during the course of a dialogue, but are still very useful to modeling the progression of the dialogue. ....

Ahrenberg, Lars, Dahlb ack, Nils, & J onsson, Arne. 1990. Discourse representation and discourse management for a natural language dialogue system. In: Proceedings of the second nordic conference on text comprehension in man and machine.

....different types. The galaxy system, is a distributed multi modal multi domain simple service system that provides users with information about for example travel and weather [ Goddeau et al. 1994; Seneff et al. 1998 ] The linlin system is also a dialogue system used for information retrieval [ Ahrenberg et al. 1990 ] which has been customised for various domains, e.g. information of second hand cars, charter trips to the Greek archipelago, and more recently timetable information for bus traffic [ Flycht Eriksson and Jonsson, 1998 ] The RailTel system [ Bennacef et al. 1996 ] and the sundial system [ ....

Lars Ahrenberg, Arne Jonsson, and Nils Dahlback. Discourse representation and discourse management for natural language interfaces. In Proceedings of the Second Nordic Conference on Text Comprehension in Man and Machine, Taby, Sweden, 1990.

.... Local Dialogue Structure There have been several proposals for the kind of dialogue unit represented in (1) using structural terms like adjacency pair [ Schegloff and Sacks, 1973 ] exchange [ Sinclair and Coulthard, 1975 ] game [ Severinson Eklundh, 1983, Carletta et al. 1997 ] IR unit [ Ahrenberg et al. 1990 ] and argumentation act [ Traum and Hinkelman, 1992 ] At an abstract level, we need a unit which can contain three moves or acts, as indicated in (2) 2) 1] Initiative: Request(Act) Instruct] 2] Response: Do(Act) 3] Feedback: Eval [ Counter Request(Act ) There are several ways in which ....

Lars Ahrenberg, Nils Dahlback, and Arne Jonsson. Discourse representation and discourse management for a natural language dialogue system. In Proceedings of the Second Nordic Conference on Text Comprehension in Man and Machine, 1990.

....into Danish. The PaTrans grammar has not been described in detail, but it seems to build on the Eurotra E Framework unification based formalism [Bech Nygaard 1988] There has also been some work, for example, in Bergen, Norway [Dyvik 1990, Dyvik 1995] and Linkoping, Sweden [Ahrenberg 1985, Ahrenberg 1990, Ahrenberg 1993] within the more up to date Lexical Functional Grammar (LFG, Section 3.2.3) and PATRII (Section 3.2.5) approaches. The Bergen grammar has composed part of a machine translation system called PONS, while the Linkoping Natural Language Interface (LINLIN) aimed at the task of ....

....up to date Lexical Functional Grammar (LFG, Section 3.2.3) and PATRII (Section 3.2.5) approaches. The Bergen grammar has composed part of a machine translation system called PONS, while the Linkoping Natural Language Interface (LINLIN) aimed at the task of language query interfacing to databases [Ahrenberg et al. 1990]. In a set of Prolog based experiments at Lund University at the end of the 80 s, a small machine translation system called SWETRA [Sigurd et al. 1992] was built using Bengt Sigurd s Reference Grammar [Sigurd 1988] a version of DCG (Section 3.2.5) with some influences from LFG. 3 Some attempts ....

[Article contains additional citation context not shown here]

Lars Ahrenberg, Arne Jonsson, and Nils Dahlback. "Discourse Representation and Discourse Management for Natural Language Interfaces ". In O. Dahl and K. Fraurud, editors, Papers from the 2nd Nordic Conference on Text Comprehension in Man and Machine, pp. 1--14, Taby --- Stockholm, Sweden, 1990.

.... go further by using a natural language interface system which is based on modern linguistic theories, which use an object oriented knowledge representation language throughout and which is capable of participating in a coherent dialogue with the user, LINLIN (LInk ping Natural Language INterface) (Ahrenberg, J nsson Dahlb ck, 1990). In LINLIN the interface is customized to a certain application using a process inspired by the method of user derived interfaces. The paper will focus on the dialogue management part of the interface. 2 LINLIN In LINLIN linguistic knowledge and domain knowledge are integrated in the same ....

....in J nsson (1991) but some of its distinguishing features needs to be presented before we can discuss how the DM is customized. Reichman (1985) describes a discourse grammar based on the assumption that a conversation can be described using conventionalized discourse rules. We argue similarly (Ahrenberg, J nsson Dahlb ck, 1990) that a segment structure inferred from the structure and content of the utterances is primary when describing a dialogue. This proposal has been criticized by for instance Levinson (1981) as not accurately describing a naturally occurring discourse, but for a restricted sublanguage in a limited ....

[Article contains additional citation context not shown here]

Ahrenberg, L., Jönsson, A. & Dahlbäck, N. (1990) Discourse Representation and Discourse Management for a Natural Language Dialogue System, To appear in Proceedings of the Second Nordic Conference on Text Comprehension in Man and Machine, Täby, Stockholm.

....met the project criteria of robustness, usefulness and cooperativeness, cf. Black et al. 1991. The dialogue model to be used as a starting point in this project was developed at the Laboratory of Natural Language Processing at Department of Computer and Information Science, Linkoping University (Ahrenberg et al. 1990, Jonsson, 1997) It is based on a number of empirical investigations into human computer natural language interaction carried out in interdisciplinary cooperations involving computer scientists, linguists and cognitive psychologists. The dialogue model does not aim to mimic human interaction for ....

Ahrenberg, L., Dahlback, N. and Jonsson, A. (1990) Discourse Representation and Discourse Management for a Natural Language Dialogue System, in Proceedings of the Second Nordic Conference on Text Comprehension in Man and Machine.

....also an intermediate analysis level, characterized by having a common topic, e.g. an object whose properties are discussed over a sequence of segments. This is, however, not used in the present model. The reason for this is that we found in our empirical analysis the notion of topic problematic (Ahrenberg, Jonsson, Dahlback, 1990) by being too simple. A sequence of segments may be connected in a number of different ways, e.g. by dealing with one object for which different properties are at issue. But it may also be the other way around, so that the same property is topical, while different objects are talked about. ....

Ahrenberg, L., Jonsson, A., & Dahlback, N. (1990). Discourse representation and discourse management for natural language interfaces. In Proceedings of the Second Nordic Conference on Text Comprehension in Man and Machine, Taby, Sweden.

....[4, 7] Furthermore, the dialogue manager should be designed to facilitate customization to various applications. 2. THE DIALOGUE MANAGER The Dialogue Manager presented in this paper is a joint effort of the members of the Natural Language Processing Laboratory at Linkoping University, Sweden [3]. It was initially designed for written interaction from an analysis of a corpus of 21 dialogues, collected in Wizard of Oz experiments using five different background systems. The Dialogue Manager was developed for written interaction to Simple Service Systems [9] Such systems require in essence ....

....of exchanges. However, a sequence of segments may be connected in a number of different ways; e.g. by dealing with one object for which different properties are at issue. But it may also be the other way around, so that the same property is topical, while different objects are talked about [3]. To specify the functional role of a move, we use the parameters Type and Topic. Type corresponds to the illocutionary type of the move. In simple service systems two sub goals can be identified [9, p. 266] 1) specify a parameter to the system and 2) obtain the specification of a parameter . ....

Lars Ahrenberg, Arne Jonsson, and Nils Dahlback. Discourse representation and discourse management for natural language interfaces. In Proceedings of the Second Nordic Conference on Text Comprehension in Man and Machine, Taby, Sweden, 1990.

....from a more general perspective and propose three design principles relating to the notion of sublanguage, the quantity of information, and asymmetries between users and systems. Customizing Interaction for Natural Language Interfaces 2 The linlin model The natural language interface linlin (Ahrenberg, Jonsson, Dahlback, 1990; Jonsson, 1991, 1993a) is designed to facilitate customization to various applications. Dialogue in linlin is modelled using dialogue objects which represent speech acts and speech act sequences. The dialogue objects are structured in terms of parameters that represent their properties and ....

Ahrenberg, L., Jonsson, A., & Dahlback, N. (1990). Discourse representation and discourse management for natural language interfaces. In Proceedings of the Second Nordic Conference on Text Comprehension in Man and machine, Taby-Stockholm, Sweden, April 25-27, 1990.

....manager directs a natural language interface and holds information needed by the modules in the interface, including the dialogue manager itself. The Dialogue Manager considered in this paper was designed from an analysis of a corpus of 21 dialogues, using five different background systems [ Ahrenberg et al. 1990; Jonsson, 1991 ] collected in Wizard Oz experiments [ Dahlback et al. 1993 ] The Dialogue Manager need to be customized to account for the sublanguage carried out in a specific application. Customization allows us to adapt the behaviour of the interface to the requirements of the application ....

....in figure 1, a sequence of segments may hang together in a number of different ways; e.g. by being about one object for which different properties are at issue. But it may also be the other way around, so that the same property is topical, while different objects are talked about (cf. Ahrenberg et al. 1990 ] Thus, only one discourse segment category is distinguished and an Initiative response (IR) structure is assumed (cf. adjacency pairs [ Schegloff and Sacks, 1973 ] where an initiative opens a segment by introducing a new goal and the response closes the segment [ Dahlback, 1991b ] To ....

Lars Ahrenberg, Arne Jonsson, and Nils Dahlback. Discourse representation and discourse management for natural language interfaces. In Proceedings of the Second Nordic Conference on Text Comprehension in Man and Machine, Taby, Sweden, 1990.

....in the dialogue objects reflect the information needed by the various processes accessing information stored in the dialogue tree. A dialogue object consists of a set of parameters for specifying the initiator, responder, context etc. needed 1 For further details of the Dialogue Manager, see (Ahrenberg et al. 1990); Jonsson, 1991) and (Jonsson, 1993) 2 The use of three categories for hierarchically structuring the dialogue is motivated from the analysis of the corpora. However, there is no claim that they are applicable to all types of dialogue, and even less so, to any type of discourse. When a ....

Ahrenberg, Lars; Jonsson, Arne; and Dahlback, Nils 1990. Discourse representation and discourse management for natural language interfaces. In Proceedings of the Second Nordic Conference on Text Comprehension in Man and machine, Taby.

No context found.

Lars Ahrenberg, Arne J onsson, and Nils Dahlb ack. Discourse representation and discourse management for a natural language dialogue system. Technical report, Institutionen f or Datavetenskap, Universitetet och Tekniska H ogskolan Link oping, 1991.

No context found.

Lars Ahrenberg, Nils Dahlback, and Arne Jonsson. 1990. Discourse representation and discourse management for a natural language dialogue system.

No context found.

Lars Ahrenberg, Arne Jonsson, and Nils Dahlback. "Discourse Representation and Discourse Management for Natural Language Interfaces ". In O. Dahl and K. Fraurud, editors, Papers from the 2nd Nordic Conference on Text Comprehension in Man and Machine, pp. 1--14, Taby --- Stockholm, Sweden, 1990.

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