We report on the strategies in automatic summarization of spontaneous spoken dialogues. Our dialogue summarization system, MIMI, recognizes key linguisitc patterns and merges information to construct a summary of conference room scheduling dialogues. Dialogues about single reservations can be accurately summarized with a simple merging scheme, but we need several extensions and changes to cope with the aboutness complexity in unrestricted dialogues --- knowing exactly how many reservations are being discussed and which reservation values must be updated by subsequent utterances. A side effect of these extensions is a new problem of overrecognition caused by spoken language disfluencies. Keywords: Spontaneous Dialogues, Summarization, Information Extraction 1. INTRODUCTION A full understanding of spoken dialogues would need a full discourse understanding plus robustness with disfluencies. Such a system would require a lot of linguistic and world knowledge, and, if at all possible, wou...

We present a semantic and pragmatic account of the anaphoric properties of past and perfect that improves on previous work by integrating discourse structure, aspectual type, surface structure and commonsense knowledge. A novel aspect of our account is that we distinguish between two kinds of temporal intervals in the interpretation of temporal operators --- discourse reference intervals and event intervals. This distinction makes it possible to develop an analogy between centering and temporal centering, which operates on discourse reference intervals. Our temporal property-sharing principle is a defeasible inference rule on the logical form. Along with lexical and causal reasoning, it plays a role in incrementally resolving underspecified aspects of the event structure representation of an utterance against the current context.

Languages differ in the concepts and real-world entities for which they have words and grammatical constructs. Therefore translation must sometimes be a matter of approximating the meaning of a source language text rather than finding an exact counterpart in the target language. We propose a translation framework based on Situation Theory. The basic ingredients are an information lattice, a representation scheme for utterances embedded in contexts, and a mismatch resolution scheme defined in terms of information flow. We motivate our approach with examples of translation between English and Japanese.

We describe a prototype SHARED GRAMMAR for the syntax of simple nominal expressions in Arabic, English, French, German, and Japanese implemented at MCC. In this grammar, a complex inheritance lattice of shared grammatical templates provides parts that each language can put together to form language-specific grammatical templates. We conclude that grammar sharing is not only possible but also desirable. It forces us to reveal crosslinguistically invariant grammatical primitives that may otherwise remain conflated with other primitives if we deal only with a single language or language type. We call this the process of GRAMMATICAL ATOMIZATION. The specific implementation reported here uses categorial unification grammar. The topics include the mono-level nominal category N, the functional distinction between ARGUMENT and NON-ARGUMENT of nominals, grammatical agreement, and word order types.

A constraint is proposed in the Centering approach to pronoun resolution in discourse. This "property-sharing" constraint requires that two pronominal expressions that retain the same Cb across adjacent utterances share a certain common grammatical property. This property is expressed along the dimension of the grammatical function SUBJECT for both Japanese and English discourses, where different pronominal forms are primarily used to realize the Cb. It is the zero pronominal in Japanese, and the (unstressed) overt pronoun in English. The resulting constraint complements the original Centering, accounting for its apparent violations and providing a solution to the interpretation of multi-pronominal utterances. It also provides an alternative account of anaphora interpretation that appears to be due to structural parallelism. This reconciliation of centering/focusing and parallelism is a major advantage. I will then add another dimension called the "speaker identification" to the constraint to handle a group of special cases in Japanese discourse. It indicates a close association between centering and th speaker's viewpoint, and sheds light on what underlies the effect of perception reports on pronoun resolution in general. These results, by drawing on facts in two very different languages, demonstrate the cross-linguistic applicability of the centering framework.