This paper considers the problem of automatic assessment of local coherence. We present a novel entity-based representation of discourse which is inspired by Centering Theory and can be computed automatically from raw text. We view coherence assessment as a ranking learning problem and show that the proposed discourse representation supports the effective learning of a ranking function. Our experiments demonstrate that the induced model achieves significantly higher accuracy than a state-of-the-art coherence model. 1

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.

This paper presents research that connects the cohesion structure of a text to the derivation of its coherence structure. Two different algorithms that derive the cohesion structure in the form of lexical paths from large thesauri are illustrated. Their results are correlated with (1) cue phrases of discourse usage and (2) coherence constraints empirically derived. A novel model of coherence structure is devised, based on the data provided by lexical paths from real world texts. 1 Introduction The inference of the coherence structure of a text is a major natural language processing endeavor, that has been stimulating a lot of theoretical work in the 80s, and more recently, promising computational research. It has been shown that coherence-based approaches increase the accuracy of some of the most important present applications of the language technology: information extraction and textual summarization. Moreover, coherence information is an important factor in building content-based W...

This article proposes a novel technique to generate natural language descriptions for a wide class of relational database queries. The approach to describing queries is phrasal and is restricted to a class of queries that return only whole schema tuples as answers. Query containment and equivalence are decidable for this class and this property is exploited in the maintenance and use of a phrasal lexicon. The query description mechanism is implemented within the STEP (Schema Tuple Query Processor) system

A reader's perception of even an "objective" text is to some degree subjective. We present the results of a pilot study in which we looked at the degree of subjectivity in readers' perceptions of lexical semantic relations, which are the building blocks of the lexical chains used in many applications in natural language processing. An example is presented in which the subjectivity reflects the reader's attitude.

The paper presents a new approach to identi-fying discourse relations, which makes use of a particular sampling method called committee-based sampling (CBS). In the committee-based sampling, multiple learning models are gener-ated to measure the utility of an input example in classification; if it is judged as not useful, then the example will be ignored. The method has the effect of reducing the amount of data required for training. In the paper, we extend CBS for decision tree classifiers. With an addi-tional extension called error feedback, it is found that the method achieves an increased accuracy as well as a substantial reduction in the amount of data for training classifiers. 1

The emphasis in most process-oriented models of argumentation is placed heavily upon analysis of dialogue. The current work puts forward an account which examines the argumentation involved in persuasive monologue, drawing upon commitment-based theories of dialogue. The various differences between monologue and dialogue are discussed, with particular reference to the possibility of designing a monologue game in which commitments are dynamically incurred and updated as the monologue is created. Finally, the computational advantages of adopting such an approach are explored in the context of an existing architecture for the generation of natural language arguments. 1. Introduction + The rhetorical tradition since Aristotle has focused on the dialogical side to + There are a number of conventions adopted for this paper. The author of a monologue, whether it is written or spoken, is referred to as the `speaker', and is pronominalised using the feminine `her' when alternative phrasing w...

semantic analyses, underspecified semantic representation

This paper 1 presents a computational method for the recognition of the cohesive and coherence structures of texts. A large lexical knowledge base built on top of WordNet provides with the lexico-semantic information that needs to be mined. A path-finding algorithm returns the cohesive structure of a text with results that outperform previous approaches. The lexical paths contained in the cohesive structures are used to (1) build patterns of association between cue phrases and coherence relations and (2) to find the lexical characteristics of coherence categories. Finally, the textual coherence structure is recognized by giving priority to the coherence constrains induced by cue phrases. The paper presents also the performance of building the coherence structure for several texts. Introduction In a text, a sequence of sentences tends to convey information about a certain topic, and by doing so, they use related words, providing the text with the quality of unity. This property of se...

. Argument represents an opportunity for a system to convince a possibly sceptical or resistant audience of the veracity of its own beliefs. This ability is a vital component of rich communication, facilitating explanation, instruction, cooperation and conflict resolution. In this paper, a proposal is presented for the architecture of a system capable of constructing arguments which comprise not only logical structure but also the rhetoric which makes that structure persuasive. The design of the architecture has made use of the wealth of naturally occurring argument, which, unlike much natural language, is particularly suited to analysis due to its clear aims and structure. The proposed framework is based upon a core hierarchical planner conceptually split into four levels of processing, the highest being responsible for abstract, intentional and pragmatic guidance, and the lowest handling realisation into natural language. The higher levels will have control over both the logical form...