JANUS is a multi-lingual speech-tospeech translation systern designed to facilitate communication between two parties engaged in a spontaneous con- versation in a limitcd dmnain. In an attempt to achieve both robustness and translation accuracy we use two different translation components: the GLI{ module, designed to be more accurate, and the Phoenix module, designed to be more robust. We analyze the strengths and weaknesses of each of thc approaches and describe our work on combining them. Another recent focus has been on developing a detailed endto -end evaluation procedure to measure the performance and effectiveness of the system. We present our most recent Spanish-to-English performance cvalua- tion results.

As part of the JANUS speech-to-speech translation project[5], we have developed a translation system that successfully parses full utterances and is effective in parsing spontaneous speech, which is often syntactically ill-formed. The system is concept-based, meaning that it has no explicit notion of a sentence but rather views each input utterance as a potential sequence of concepts. Generation is performed by translating each of these concepts in whole phrases into the target language, consulting lookup tables only for low-level concepts such as numbers. Currently, we are working on an appointment scheduling task, parsing English, German, Spanish, and Korean input and producing output in those same languages and also Japanese. 1

As part of the JANUS speech-to-speech translation project, we have developed a robust translation system based on the information structures inherent to the task being performed. The basic premise is that the structure of the information to be transmitted is largely independent of the language used to encode it. Our system performs no syntactic analysis; speaker utterances are parsed into semantic chunks, which can be strung together without grammatical rules, and passed through a simple template-based translation module. We have achieved encouraging coverage rates on English, German and Spanish input with English, German and Spanish output.

This paper describes JANUS-III, our most recent version of the JANUS speech-to-speech translation system. We present anoverview of the system and focus on how system design facilitates speech translation between multiple languages, and allows for easy adaptation to new source and target languages. We also describe our methodology for evaluation of end-to-end system performance with a variety of source and target languages. For system development and evaluation, we haveexperimented with both push-to-talk as well as cross-talk recording conditions. To date, our system has achieved performance levels of over 80 % acceptable translations on transcribed input, and over 70 % acceptable translations on speech input recognized with a 75-90 % word accuracy. Our current major research is concentrated on enhancing the capabilities of the system to deal with input in broad and general domains. 1.

ing . . . . . . . . . . . . . . . . . . . . 20 6.2 Written Text Summarization . . . . . . . . . . . . . . . . . . . . 20 6.2.1 Systems Using Term Statistics (and Heuristics) . . . . . . 20 6.2.2 A System Using Text Understanding . . . . . . . . . . . . 23 6.2.3 An Abstracting Evaluation Study . . . . . . . . . . . . . . 25 6.3 Spoken Language Summarization . . . . . . . . . . . . . . . . . . 25 7 Conclusions and Future Perspectives 26 8 References 27 1 1 Introduction While the science of automatic information retrieval (IR) has been around for quite a long time now --- basically, as long as databases of documents have been existing ---, the scientific community only quite recently has been moving to the field of automatic information extraction (IE), maybe for as long as a decade now. There are at least two reasons for that: first, the amount of on-line (and off-line) textual data has been increasing exponentially, without a limit in sight (keywords are electronic newswires, World Wi...

Attempts at discourse processing of spontaneously spoken dialogue face several difficulties: multiple hypotheses that result from the parser’s attempts to make sense of the output from the speech recognizer, ambiguity that results from segmentation of multi-sentence utterances, and cumulative error — errors in the discourse context which cause further errors when subsequent sentences are processed. In this paper we will describe our robust parsers, our procedures for segmenting long utterances, and two approaches to discourse processing that attempt to deal with ambiguity and cumulative error. 1.

Much work has been done in dialogue modeling for spoken and multi-modal humancomputer interaction. Problems can arise in situations that do not correspond to the dialogue model. For this reason, we propose information-centered dialogue processing in which the actions to be taken by the dialogue system are determined as a function of the information available in the discourse, the database and the domain model. In order to arrive at fully specified representations of the intended actions, the specificity of the representations is increased by unification, integrating information from multi-modal input, database access and domain knowledge. Our approach differs from other state-of-the-art systems in that it does not rely on explicit dialogue models. Instead, we show how partial and underspecified representations of the situation can be used in a spoken dialogue system to generate clarification questions and to guide the user to arrive at his or her communicative goal. We show furthermor...

As part of the JANUS speech-to-speech translation project[5], we have developed a translation system that successfully parses full utterances and is e ective in parsing spontaneous speech, which is often syntactically ill-formed. The system is concept-based, meaning that it has no explicit notion of a sentence but rather views each input utterance asapotential sequence ofconcepts. Generation is performed bytranslating each of these concepts in whole phrases into the target language, consulting lookup tables only for low-level concepts such as numbers. Currently, we are working on an appointment scheduling task, parsing English, German, Spanish, and Korean input and producing output in those same languages and also Japanese. 1.

In this paper we investigate the possibility of translating continuous spoken conversations in a cross-talk environment. This is a task known to be difficult for human translators due to several factors. It is characterized by rapid and even overlapping turn-taking, a high degree of co-articulation, and fragmentary language. We describe experiments using both push-to-talk as well as cross-talk recording conditions. Our results indicate that conversational speech recognition and translation is possible, even in a free crosstalk environment. To date, our system has achieved performances of over 80% acceptable translations on transcribed input, and over 70 % acceptable translations on speech input recognized with a 70-80 % word accuracy. The system’s performance on spontaneous conversations recorded in a cross-talk environment is shown to be as good and

We demonstrate the portability of a stochastic method for understanding natural language from a setting of human-machine interactions (ATIS - Air Travel Information Services and MASK - Multimodal Multimedia Automated Service Kiosk) into the more open one of human-to-human interactions. The application we use is the English Spontaneous Speech Task (ESST) for multilingual appointment scheduling. Spoken language systems developed for this task translate spontaneous conversational speech among different languages. 1. INTRODUCTION In this paper, a stochastic component for natural language understanding, initially developed as a part of a spoken language system for the information retrieval applications ATIS (Air Travel Information Services) and MASK (Multimodal Multimedia Automated Service Kiosk) [7], is ported to a multilingual, appointment scheduling task, the English SpontaneousSpeechTask(ESST) [8]. Machine translation systems combine speech recognition, natural language understanding ...