this paper, we focus on two levels of analysis. The first involves the relation between the information conveyed in consecutive elements of a coherent discourse. Thus, for example, one utterance may describe an event that can be presumed to be the cause of another event described in the subsequent utterance. This causal relation is at what we will call the informational level. The second level of relation results from the fact that discourses are produced to effect changes in the mental state of the discourse participants. In coherent discourse, a speaker is carrying out a consistent plan to achieve the intended changes, and consecutive discourse elements are related to one another by means of the ways in which they participate in that plan. Thus, one utterance may be intended to increase the likelihood that the hearer will come to * Department of Computer Science and Intelligent Systems Program, University of Pittsburgh, Pittsburgh, PA 15260. e-mail: jmoore,pollackcs.pitt.edu 1 In addition, intentional structure is needed to make certain types of choices during the generation process, e.g., how to refer to an object (Appelt 1985)

The notion that a text is coherent in virtue of the `relations' that hold between its component spans currently forms the basis for an active research programme in discourse linguistics. Coherence relations feature prominently in many theories of discourse structure, and have recently been used with considerable success in text generation systems. However, while the concept of coherence relations is now common currency for discourse theorists, there remains much confusion about them, and no standard set of relations has yet emerged. The aim of this thesis is to contribute towards the development of a standard set of relations. We begin from an explicitly empirical conception of relations: they are taken to model a collection of psychological mechanisms operative during the tasks of reading and writing. This conception is fleshed out with reference to psychological theories of skilled task performance, and to Rosch's notion of the basic level of categorisation. A methodology for investi...

This thesis is an inquiry into the nature of the high-level, rhetorical structure of unrestricted natural language texts, computational means to enable its derivation, and two applications (in automatic summarization and natural language generation) that follow from the ability to build such structures automatically. The thesis proposes a first-order formalization of the high-level, rhetorical structure of text. The formalization assumes that text can be sequenced into elementary units; that discourse relations hold between textual units of various sizes; that some textual units are more important to the writer's purpose than others; and that trees are a good approximation of the abstract structure of text. The formalization also introduces a linguistically motivated compositionality criterion, which is shown to hold for the text structures that are valid. The thesis proposes, analyzes theoretically, and compares empirically four algorithms for determining the valid text structures of ...

The problem of organizing information for multidocument summarization so that the generated summary is coherent has received relatively little attention. While sentence ordering for single document summarization can be determined from the ordering of sentences in the input article, this is not the case for multidocument summarization where summary sentences may be drawn from different input articles. In this paper, we propose a methodology for studying the properties of ordering information in the news genre and describe experiments done on a corpus of multiple acceptable orderings we developed for the task. Based on these experiments, we implemented a strategy for ordering information that combines constraints from chronological order of events and topical relatedness. Evaluation of our augmented algorithm shows a significant improvement of the ordering over two baseline strategies. 1.

We describe our experience in developing a discourse-annotated corpus for community-wide use. Working in

To explain complex phenomena, an explanation system must be able to select information from a formal representation of domain knowledge, organize the selected information into multisentential discourse plans, and realize the discourse plans in text. Although recent years have witnessed significant progress in the development of sophisticated computational mechanisms for explanation, empirical results have been limited. This paper reports on a seven-year effort to empirically study explanation generation from semantically rich, large-scale knowledge bases. In particular, it describes KNIGHT, a robust explanation system that constructs multisentential and multiparagraph explanations from the Biology Knowledge Base, a large-scale knowledge base in the domain of botanical anatomy, physiology, and development. We introduce the Two-Panel evaluation methodology and describe how KNIGHT'S performance was assessed with this methodology in the most extensive empirical evaluation conducted on an explanation system. In this evaluation, KNIGHT scored within "half a grade " of domain experts, and its performance exceeded that of one of the domain experts. 1.

Abstract not found

... In this paper, we argue that the main theories representing these two approaches, RST (Mann and Thompson 1988) and G&S (Grosz and Sidner 1986), make similar claims about how speakers' intentions determine a structure of their discourse. The similarity occurs because the nucleus-satellite relation among text spans in RST corresponds to the dominance relation among intentions in G&S. Building on this similarity, we sketch a partial mapping between the two theories to show that the main points of the two theories are equivalent. Furthermore, the additional claims found in only RST or only G&S are largely consistent. The issue of what structure is determined by semantic (domain) relations in the discourse and how this structure might be related to the intentional structure is discussed. We suggest the synthesis of the two theories would be useful to researchers in both natural language interpretation and generation.

This paper introduces Linguistic Style Improvisation, a theory and set of algorithms for improvisation of spoken utterances by artificial agents, with apphcations to interactive story and dialogue systems. We argue that hnguistic style is a key aspect of character, and show how speech act representations common in AI can provide abstract representations from which computer characters can improvise. We show that the mechanisms proposed introduce the possibility of socially oriented agents, meet the requirements that lifehke characters be behevable, and satisfy particular criteria for improvisation proposed by Hayes-Roth.

This paper presents a plan-based architecture for response generation in collaborative consultation dialogues, with emphasis on cases in which the system (consultant) and user (executing agent) disagree. Our work contributes to an overall system for collaborative problem-solving by providing a plan-based framework that captures the Propose-Evaluate-Modify cycle of collaboration, and by allowing the system to initiate subdialogues to negotiate proposed additions to the shared plan and to provide support for its claims. In addition, our system handles in a unified manner the negotiation of proposed domain actions, proposed problem-solving actions, and beliefs proposed by discourse actions. Furthermore, it captures cooperative responses within the collaborative framework and accounts for why questions are sometimes never answered.