Mats Wirén

The Spoken Language Translator

This original volume describes the Spoken Language Translator (SLT), one of the first major automatic speech translation projects. The SLT system can translate between English, French, and Swedish in the domain of air travel planning, using a vocabulary of about 1500 words, and with an accuracy of about 75%. The authors detail the language processing components, largely built on top of the SRI Core Language Engine, using a combination of general grammars and techniques that allow them to be rapidly customized to specific domains. They base speech recognition on Hidden Markov Mode technology, and use versions of the SRI DECIPHER system. This account of SLT is an essential resource for researchers interested in knowing what is achievable in spoken-language translation today.

Robust spoken language understanding in a computer game

We present and evaluate a robust method for the interpretation of spoken input to a conversational computer game. The scenario- of the game is that of a player interacting with embodied fairy-tale ...

Improving Readability of Swedish Electronic Health Records through Lexical Simplification: First Results

This paper describes part of an ongoing effort to improve the readability of Swedish electronic health records (EHRs). An EHR contains systematic documentation of a single patient’s medical history across time, entered by healthcare professionals with the purpose of enabling safe and informed care. Linguistically, medical records exemplify a highly specialised domain, which can be superficially characterised as having telegraphic sentences involving displaced or missing words, abundant abbreviations, spelling variations including misspellings, and terminology. We report results on lexical simplification of Swedish EHRs, by which we mean detecting the unknown, out-ofdictionary words and trying to resolve them either as compounded known words, abbreviations or misspellings.

Multi-slot semantics for natural-language call routing systems

Statistical classification techniques for natural-language call routing systems have matured to the point where it is possible to distinguish between several hundreds of semantic categories with an accuracy that is sufficient for commercial deployments. For category sets of this size, the problem of maintaining consistency among manually tagged utterances becomes limiting, as lack of consistency in the training data will degrade performance of the classifier. It is thus essential that the set of categories be structured in a way that alleviates this problem, and enables consistency to be preserved as the domain keeps changing. In this paper, we describe our experiences of using a two-level multi-slot semantics as a way of meeting this problem. Furthermore, we explore the ramifications of the approach with respect to classification, evaluation and dialogue design for call routing systems.

Experiences of an In-Service Wizard-of-Oz Data Collection for the Deployment of a Call-Routing Application

This paper describes our experiences of collecting a corpus of 42,000 dialogues for a call-routing application using a Wizard-of-Oz approach. Contrary to common practice in the industry, we did not use the kind of automated application that elicits some speech from the customers and then sends all of them to the same destination, such as the existing touch-tone menu, without paying attention to what they have said. Contrary to the traditional Wizard-of-Oz paradigm, our data-collection application was fully integrated within an existing service, replacing the existing touch-tone navigation system with a simulated call-routing system. Thus, the subjects were real customers calling about real tasks, and the wizards were service agents from our customer care. We provide a detailed exposition of the data collection as such and the application used, and compare our approach to methods previously used.

Swedification patterns of Latin and Greek affixes in clinical text

Swedish medical language is rich with Latin and Greek terminology which has undergone a Swedification since the 1980s. However, many original expressions are still used by clinical professionals. T ...

Robust parsing and spoken negotiative dialogue with databases

This paper presents a robust parsing algorithm and semantic formalism for the interpretation of utterances in spoken negotiative dialogue with databases. The algorithm works in two passes: a domain-specific pattern-matching phase and a domain-independent semantic analysis phase. Robustness is achieved by limiting the set of representable utterance types to an empirically motivated subclass which is more expressive than propositional slot–value lists, but much less expressive than first-order logic. Our evaluation shows that in actual practice the vast majority of utterances that occur can be handled, and that the parsing algorithm is highly efficient and accurate.

Modelling the informativeness and timing of non-verbal cues in parent–child interaction

How do infants learn the meanings of their first words? This study investigates the informativeness and temporal dynamics of non-verbal cues that signal the speakers referent in a model of early w ...

Longitudinal Studies of Variation Sets in Child-directed Speech

One of the characteristics of child-directed speech is its high degree of repetitiousness. Sequences of repetitious utterances with a constant intention, variation sets, have been shown to be corre ...