We describe a novel method for coping with ungrammatical input based on the use of chart-like data structures, which permit anytime processing. Priority is given to deep syntactic analysis. Should this fail, the best partial analyses are selected, according to a shortest-paths algorithm, and assembled in a robust processing phase. The method has been applied in a speech translation project with large HPSG grammars.

This paper gives an overview of the recent improvements in dynamic programming search for large vocabulary continuous speech recognition: search using lexical trees, time-conditioned search and word graph construction.

In 1995 our department was involved in two projects in the field of continuous speech recognition. The main aim of these two strongly related projects was the development of basic technology that can be used to build advanced telephone-based systems for providing information about public transport. A short description of the work carried out within these projects is provided in the present article. 1. Introduction During the last decade the performance of spoken dialogue systems has improved substantially. At the moment, the quality of these systems seems to be able to support a number of simple practical tasks in small and clearly delimited domains. As a result, much effort is spent nowadays to develop prototype telephone-based information systems in different countries. These systems are reminiscent of the well-known Air Travel Information System (ATIS) task that has been a focal point in the American ARPA-project. In Europe two MLAP (Multi-Lingual Action Plan) projects concerning p...

Abstract not found

Abstract. We are going to present an architecture for natural language processing sys-tems especially designed for spontaneous speech. We introduce the notion of a layered chart for information separation and information hiding. Complex distributed systems can be built using this approach which do not rely on a central data structure or control model. We describe a typed feature formalism with appropriateness suitable for such sys-tems by providing a compact and relocatable storage schema for feature structures. We set out an architecture for spontaneous speech translation that relies on partial parsing to minimize complex operations at an early syntactic stage, but rather carries out ut-terance integration after building up small scale syntactic objects. We adopt the notion of a variable depth of analysis by providing a module specialized to the translation of idioms and show how an global scoring schema for sub-paths in graphs can be used to block further operation on idioms already accounted for. 1

The control of continuous speech recognition by a context-free based language model requires a parsing process which may overload the acoustic decoding algorithm. We present a new approach to integrate such a language model in the search process. This approach extends the beam search Viterbi algorithm. In our case, the pruning technique not only selects the most likely acoustic hypotheses but also governs the dynamic expansion of a network structure. This algorithm is general enough to cope with the self-embedded recursivity of context-free languages and it favourably compares with other parsing techniques applied to spoken inputs. We present results which show that the syntactical knowledge may be efficiently included at the frame level of an acoustic decoding algorithm. Keywords: Continuous Speech Recognition, Context-Free Language Models, Beam Search Viterbi Algorithm. 1. INTRODUCTION Many continuous speech understanding systems use different language models for recognition, on the ...

We discuss the issue of how a speech understanding system can be made robust against spontaneous speech phenomena and speech recognition errors. We argue that treating these problems before or during parsing raises certain problems which can be circumvented by an approach handling them after parsing, and we describe how this approach is currently being investigated in the Verbmobil speech translation project. This paper will appear in the proceedings of ECAI 1998 (Worm and Rupp 1998). Contents 1 INTRODUCTION 2 2 THE VERBMOBIL PROJECT 2 2.1 The system architecture . . . . . . . . . . . . . . . . . . . . . . . . . 2 2.2 The VIT format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 PROBLEMS IN SPEECH PROCESSING 3 3.1 Spontaneous speech phenomena . . . . . . . . . . . . . . . . . . . . 3 3.2 Speech recognition errors . . . . . . . . . . . . . . . . . . . . . . . . 4 4 APPROACHES TO ROBUST SPEECH PROCESSING 4 4.1 Before parsing . . . . . . . . . . . . . . . . . . . . . ...

In this paper we present a contextual extension to ONTOSCORE, a system for scoring sets of concepts on the basis of an ontology. We apply the contextually enhanced system to the task of scoring alternative speech recognition hypotheses (SRH) in terms of their semantic coherence. We conducted

Large vocabulary speech recognition applications can benefit from an efficient data structure for representing large numbers of acoustic hypotheses compactly. Word graphs or lattices generated by acoustic recognition engines are generally not compact and must be post-processed to keep lattice sizes small; however, algorithms designed for this task need to reduce the size of the lattice without either eliminating hypotheses or distorting their relative acoustic probabilities. In this paper, we will discuss the relevant criteria for measuring graph size, compare the advantages of two different structures for graphs, and introduce a new data structure and compression algorithm which give additional graph compression and maintain exact hypothesis path scores by storing probability information on both nodes and arcs within the graph. 1. INTRODUCTION Many recognition systems use word lattices or graphs as mechanisms for representing sentence hypotheses and interfacing with additional knowle...

Modern AI systems need theoretically sound, easy to use, communication models in order to be able to explore distributed computing and agent-oriented operation. ICE (Intarc Communication Environment) represents such a system. It grounds on the theoretical framework of CSP (Communicating Sequential Processes) and implements channels as bidirectional, asynchronous data streams that can be configured in various ways. On top of PVM (Parallel Virtual Machine), a defacto standard of message passing systems, software layers have been built that implement the channel operation modes together with interfaces for programming languages most often used in the AI community. A separate layer supports the use of complex data types, as they often arise in speech processing. We describe the design of ICE with a focus on configuration and synchronization during the creation of channels. Keywords: Parallel/Distributed applications: Computer natural language processing; Heterogeneous software systems; So...