In this paper, we investigate two statistical methods for spoken language understanding based on statistical machine translation. The first approach employs the source-channel paradigm, whereas the other uses the maximum entropy framework. Starting with an annotated corpus, we describe the problem of natural language understanding as a translation from a source sentence to a formal language target sentence. We analyze the quality of different alignment models and feature functions and show that the direct maximum entropy approach outperforms the source channel-based method. Furthermore, we investigate how both methods perform if the input sentences contain speech recognition errors. Finally, we investigate a new approach to combine speech recognition and spoken language understanding. For this purpose, we employ minimum error rate training which directly optimizes the final evaluation criterion. By combining all knowledge sources in a log-linear way, we show that we can decrease both the word error rate and the slot error rate. Experiments were carried out on two German inhouse corpora for spoken dialogue systems.

In this paper, we introduce different features for dialogue course management and investigate their effect on the system’s behaviour for choosing the subsequent dialogue action during a dialogue session. Especially, we investigate whether the system is able to detect and resolve ambiguities, and if it always chooses that state which leads as quickly as possible to a final state that presumably meets the user’s request. The criteria and used data structures are independently from the underlying domain and can therefore be used for different applications of spoken dialogue systems.

In this paper, we present several confidence measures for (statistical) machine translation. We introduce word posterior probabilities for words in the target sentence that can be determined either on a word graph or on an N best list. Two alternative confidence measures that can be calculated on N best lists are proposed. The performance of the measures is evaluated on two different translation tasks: on spontaneously spoken dialogues from the domain of appointment scheduling, and on a collection of technical manuals.

Over the past years, automatic dialogue systems and telephonebased machine inquiry systems have received increasing attention. In addition to an automatic speech recognizer and a dialogue manager, such systems consist of a natural language understanding (NLU) component. Some of the most investigated approaches to NLU are rule-based methods as Stochastic Grammars, which are often written manually. However, the sole usage of rule-based methods can turn out to be inflexible and the problem of reusability occurs. When extending the application scenario or changing the application's domain itself, a large part of the set of rules often must be rewritten. Therefore, techniques are desirable which help to reduce the manual effort when building up an NLU component for a new domain. In this paper we investigate an approach to NLU, which is derived from the field of statistical machine translation. Starting from a conceptual annotated corpus, we describe the problem of NLU as a translation from a source sentence to a formallanguage target sentence. Doing this, we will mainly focus on the quality of different alignment models between source and target sentences. Even though the usage of grammars cannot be totally avoided in NLU-systems, it is our goal to reduce their employment and learn the dependencies between words and their meaning automatically. Experiments were performed on the Philips in-house TABA corpus, which is a text corpus in the domain of a German train timetable information system. 1.

In this paper, we present several confidence measures for large vocabulary continuous speech recognition. We propose to estimate the confidence of a hypothesized word directly as its posterior probability, given all acoustic observations of the utterance. These probabilities are computed on word graphs using a forward-backward algorithm. We also study the estimation of posterior probabilities on N-best lists instead of word graphs and compare both algorithms in detail. In addition, we compare the posterior probabilities with two alternative confidence measures, i.e., the acoustic stability and the hypothesis density. We present experimental results on five different corpora: the Dutch ARISE 1k evaluation corpus, the German Verbmobil '98 7k evaluation corpus, the English North American Business '94 20k and 64k development corpora, and the English Broadcast News '96 65k evaluation corpus. We show that the posterior probabilities computed on word graphs outperform all other confidence measures. The relative reduction in confidence error rate ranges between 19% and 35% compared to the baseline confidence error rate.

In this paper we present and compare several confidence measures for large vocabulary continuous speech recognition. We show that posterior word probabilities computed on word graphs and N-best lists clearly outperform non-probabilistic confidence measures, e.g. the acoustic stability and the hypothesis density. In addition, we prove that the estimation of posterior word probabilities on word graphs yields better results than their estimation on N-best lists and discuss both methods in detail. We present experimental results on three different corpora, the English NAB '94 20k development corpus, the German VERBMOBIL '96 evaluation corpus and a Dutch corpus, which has been recorded with a train timetable information system in the ARISE project. 1. INTRODUCTION In previous studies, the combination of several confidence features was investigated. These features were collected during the acoustic decoding process, e.g. [1] or were extracted from Nbest lists and word graphs, e.g. [2, 5]. ...

Estimates of confidence for the output of a speech recognition system can be used in many practical applications of speech recognition technology. They can be employed for detecting possible errors and can help to avoid undesirable verification turns in automatic inquiry systems. We propose to estimate the confidence in a hypothesized word as its posterior probability, given all acoustic feature vectors of the speaker utterance. The basic idea of our approach is to estimate the posterior word probabilities as the sum of all word hypothesis probabilities which represent the occurrence of the same word in more or less the same segment of time. The word hypothesis probabilities are approximated by paths in a wordgraph and are computed using a simplified forward-backward algorithm. We present experimental results on the North American Business (NAB'94) and the German Verbmobil recognition task.