We focus on the production of efficient descriptions of objects, actions and events. We define a type of efficiency, textual economy, that exploits the hearer's recognition of inferential links to material elsewhere within a sentence. Textual economy leads to efficient descriptions because the material that supports such inferences has been included to satisfy independent communicative goals, and is therefore overloaded in the sense of Pollack [18]. We argue that achieving textual economy imposes strong requirements on the representation and reasoning used in generating sentences. The representation must support the generator's simultaneous consideration of syntax and semantics. Reasoningmust enable the generator to assess quickly and reliably at any stage how the hearer will interpret the current sentence, with its '-(inc6mplete)syntax and'semantics. We show that these representational and reasoning requirements are met in the SPUD system for sentence planning and realization.

We show that discourse structure need not bear the full burden of conveying discourse relations by showing that many of them can be explained nonstructurally in terms of the grounding of anaphoric presuppositions (Van der Sandt, 1992). This simplifies discourse structure, while still allowing the realisation of a full range of discourse relations. This is achieved using the same semantic machinery used in deriving clause-level semantics.

Prior research into embodied interface agents has found that users like them and find them engaging. In this paper, we argue that embodiment can serve an even stronger function if system designers use actual human conversational protocols in the design of the interface. Communicative behaviors such as salutations and farewells, conversational turn-taking with interruptions, and referring to objects using pointing gestures are examples of protocols that all native speakers of a language already know how to perform and that can thus be leveraged in an intelligent interface. We discuss how these protocols are integrated into Rea, an embodied, multi-modal conversational interface agent who acts as a real-estate salesperson, and we show why embodiment is required for their successful implementation.

The incremental algorithm introduced in (Dale and Reiter, 1995) for producing distinguishing descriptions does not always generate a minimal description. In this paper, I show that when generalised to sets of individuals and disjunctive properties, this approach might generate unnecessarily long and ambiguous and/or epistemically redundant descriptions. I then present an alternative, constraint-based algorithm and show that it builds on existing related algorithms in that (i) it produces minimal descriptions for sets of individuals using positive, negative and disjunctive properties, (ii) it straightforwardly generalises to n-ary relations and (iii) it is integrated with surface realisation.

In this paper we discuss the application of aspects of a psychological theory about the relationship between speech and gesture to the implementation of interactive dialogue systems. We first lay out some uncontroversial facts about the interaction of speech and gesture in conversation and describe some psychological theories put forth to explain those data, settling on one theory as being the most interesting for interactive dialogue systems. We then lay out our implementation of an interactive dialogue system that is informed by the theory--- concentrating on two particular claims of the theory: that gesture and speech reflect a common conceptual source; and that the content and form of gesture is tuned to the communicative context and the actor's communicative intentions. We compare our work to some other similar interactive systems, and conclude with some thoughts about how both implementation and theory can benefit from this kind of close partnership. Epigraph Pan...

In this paper we describe the computational and architectural requirements for systems which support real-time multimodal interaction with an embodied conversational character. We argue that the three primary design drivers are real-time multithreaded entrainment, processing of both interactional and propositional information, and an approach based on a functional understanding of human face-toface conversation. We then present an architecture which meets these requirements and an initial conversational character that we have developed who is capable of increasingly sophisticated multimodal input and output in a limited application domain.

the present author, have offered different but related accounts of intonation structure in English and some other languages. These accounts share the assumption that the system of tones identified by Pierrehumbert (1980), as modified by Pierrehumbert and Beckman (1988) and Silverman et al. (1992), has as transparent and type-driven a semantics in these languages as do their words and phrases. While the semantics of intonation in English concerns information structure and propositional attitude, rather than the predicate-argument relations and operator-scope relations that are familiar from standard semantics in the spirit of the papers collected as Montague 1974, this information-structural semantics is fully compositional, and can be regarded as a component of the same semantic system. The present paper builds on Steedman (1991) and Steedman (2000a) to develop a new semantics for intonation structure, which shares with the earlier versions the property of being fully integrated into Combinatory Categorial Grammar (CCG, see Steedman 2000b, hereafter SP). This grammar integrates intonation structure into surface derivational structure and the associated Montague-style compositional semantics, even when the intonation structure departs from the restrictions of traditional surface structure. Many of the diverse discourse meanings that have been attributed to intonational tunes are shown to arise via conversational implicature from more primitive literal meanings distinguished along the three dimensions of information structure, speaker/hearer commitment, and contentiousness.

People engage in small talk and conversational storytelling to provide information in engaging ways and to serve interpersonal goals such as building rapport and credibility. Embodied conversational interface agents, that represent the computer in its interaction with human users by way of a graphical humanoid body on a screen, can also profitably engage in these behaviors. We describe the ongoing development of an embodied conversational interface agent that is capable of multimodal input understanding and output generation and operates in a limited application domain in which both social and task-oriented dialogue are important. We discuss our plans for adding small talk and conversational storytelling capabilities to the system, and the unique requirements these phenomena place on the understanding, discourse planning, and generation components of a real-time conversational interface. Introduction Interlocutors typically have multiple goals in conversation (Tracy and Coupland 1991)...

I consider abstract instructions, which provide indirect descriptions of actions in cases when a speaker has key information that a hearer can use to identify the right action to perform, but the speaker alone cannot identify that action. Principled generation of abstract instructions requires a system to assess whether an instruction provides sufficient information for the user to draw appropriate inferences about action from it. I sketch a framework for specifying, computing, and accessing those assessments in natural language generation.

Abstract not found