A speech to speech translation system built from standard components

Rayner, Manny; Alshawi, Hiyan; Bretan, Ivan; Carter, David; Digalakis, Vassilios; Gambäck, Björn; Kaja, Jaan; Karlgren, Jussi; Lyberg, Bertil; Pulman, Stephen; Price, Patti; Samuelsson, Christer

doi:10.3115/1075671.1075719

Cited by 11 publications

(7 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, the specialized grammar is used to search for full parses. The scheme is fully implemented within a version of the Spoken Language Translator system (Rayner et al, 1993;Agniis et al, 1994), and is normally applied to input in the form of small lattices of hypotheses produced by a speech recognizer.…”

Section: Introductionmentioning

confidence: 99%

Fast parsing using pruning and grammar specialization

Rayner¹,

Carter²

1996

Proceedings of the 34th Annual Meeting on Association for Computational Linguistics -

Self Cite

View full text Add to dashboard Cite

We show how a general grammar may be automatically adapted for fast parsing of utterances from a specific domain by means of constituent pruning and grammar specialization based on explanation-based learning. These methods together give an order of magnitude increase in speed, and the coverage loss entailed by grammar specialization is reduced to approximately half that reported in previous work. Experiments described here suggest that the loss of coverage has been reduced to the point where it no longer causes significant performance degradation in the context of a real application.

show abstract

Section: Introductionmentioning

confidence: 99%

Fast parsing using pruning and grammar specialization

Rayner¹,

Carter²

1996

Proceedings of the 34th Annual Meeting on Association for Computational Linguistics -

Self Cite

View full text Add to dashboard Cite

show abstract

“…In these experiments the scheme was applied to the grammar of a version of the SRI Core Language Engine [Alshawi ed. 1992] adapted to the Atis domain for a speech-translation task [Rayner et al 1993] and large corpora of real user data collected using Wizardof-Oz simulation. The resulting specialized grammar was compiled into LR parsing tables, and a special LR parser exploited their special properties, see [Samuelsson 1994b].…”

Section: Introductionmentioning

confidence: 99%

Grammar specialization through entropy thresholds

Samuelsson

1994

Proceedings of the 32nd Annual Meeting on Association for Computational Linguistics -

Self Cite

View full text Add to dashboard Cite

Explanation-based generalization is used to extract a specialized grammar from the original one using a training corpus of parse trees. This allows very much faster parsing and gives a lower error rate, at the price of a small loss in coverage. Previously, it has been necessary to specify the tree-cutting criteria (or operationality criteria) manually; here they are derived automatically from the training set and the desired coverage of the specialized grammar. This is done by assigning an entropy value to each node in the parse trees and cutting in the nodes with sufficiently high entropy values.

show abstract

“…CLARE [10] -which includes the Core Language Engine [11] as a component -is a general purpose natural-language processing system developed at SRI International, Cambridge. It has been used to build database query applications [12], spoken language dialogue systems [13] and bidirectional machine translation prototypes [14]. In the configuration used here, CLARE carries out morphological, syntactic, semantic and some contextual analysis on an input sentence and then presents one or more possible logical forms representing the interpretation(s) of the sentence.…”

Section: A Window-based Interfacementioning

confidence: 99%

A Method for Controlling the Production of Specifications in Natural Language

Macías¹,

Pulman²

1995

The Computer Journal

View full text Add to dashboard Cite

Before a system can be formally defined, it is common to write a specification in a natural language as the basis for the formal definition. Natural languages are not well suited for this task; documentation written in a natural language is often ambiguous and imprecise. Inspection of real documentation also reveals that, without special training, most writers do not produce concise, clear or consistent statements. We present here an experimental interface to a general-purpose natural-language processing system designed to control the writing of specification statements in a natural language. The interface is designed to reduce the degree of imprecision and ambiguity in the natural-language statements, as well as contributing to the writing of shorter statements, with clear structure and punctuation. We compare our approach to similar work in this area. I N T R O D U C T I O NIt is generally accepted that the initial phase in the production of software must be a careful description of the system to be developed [1]. Any such specification should ideally be carried out in a formal language, but it is almost always the case that a previous specification in a natural language must be written. This is because formalspecification languages are often intelligible only to a small group of people, and formal specifications must thus be translated into a language that both client and developers can understand [2]. Several techniques have been advanced to assist non-specialists in the understanding of formal specifications, but natural-language specifications will continue to be used for the foreseeable future.The use of natural language in this context is not without problems. It is well known that natural languages lack the properties we associate with those languages used for formal specification tasks, namely well-defined syntax and semantics. In order to use natural languages for specification purposes, we must therefore carefully consider how to do it, in order to reduce the risk of producing deficient specifications. It has been noted in the literature that the requirements specification phase, particularly the production of natural-language specifications, is one of the weakest links in the process of formally developing a system (e.g. [3][4][5]).We describe in this paper a system consisting of a window-based interface to a natural-language processing system designed to write specifications in a controlled manner. Our primary goal is to provide a tool that can be used for this purpose in any setting. We assume that the typical users will be those with enough technical knowledge of the system to be specified, but not necessarily conversant with formal specification methods. Before we describe in more detail our system, we review the most salient characteristics found in some specification documents, and the problems they present. 1The fundamental problem of natural-language processing is the great amount of ambiguity at every linguistic level: morphological, syntactic, semantic and pragmatic. As expected, all these...

show abstract

A speech to speech translation system built from standard components

Cited by 11 publications

References 16 publications

Fast parsing using pruning and grammar specialization

Fast parsing using pruning and grammar specialization

Grammar specialization through entropy thresholds

A Method for Controlling the Production of Specifications in Natural Language

Contact Info

Product

Resources

About