In this paper, we describe how a research spoken dialog system was made available to the general public. The Let’s Go Public spoken dialog system provides bus schedule information to the Pittsburgh population during off-peak times. This paper describes the changes necessary to make the system usable for the general public and presents analysis of the calls and strategies we have used to ensure high performance. 1.

In this paper, a discourse modeller for conversational spoken dialogue systems, called GALATEA, is presented. Apart from handling the resolution of ellipses and anaphora, it tracks the “grounding status ” of concepts that are mentioned during the discourse, i.e. information about who said what when. This grounding information also contains concept confidence scores that are derived from the speech recogniser word confidence scores. The discourse model may then be used for concept-level error handling, i.e. grounding of concepts, fragmentary clarification requests, and detection of erroneous concepts in the model at later stages in the dialogue. 1

This article focuses on the analysis and prediction of corrections, defined as turns where a user tries to correct a prior error made by a spoken dialogue system. We describe our labeling procedure of various corrections types and statistical analyses of their features in a corpus collected from a train information spoken dialogue system. We then present results of machinelearning experiments designed to identify user corrections of speech recognition errors. We investigate the predictive power of features automatically computable from the prosody of the turn, the speech recognition process, experimental conditions, and the dialogue history. Our best-performing features reduce classification error from baselines of 25.70–28.99 % to 15.72%. 1.

Speech recognition problems are a reality in current spoken dialogue systems. In order to better understand these phenomena, we study dependencies between speech recognition problems and several higher level dialogue factors that define our notion of student state: frustration/anger, certainty and correctness. We apply Chi Square (χ2) analysis to a corpus of speech-based computer tutoring dialogues to discover these dependencies both within and across turns. Significant dependencies are combined to produce interesting insights regarding speech recognition problems and to propose new strategies for handling these problems. We also find that tutoring, as a new domain for speech applications, exhibits interesting tradeoffs and new factors to consider for spoken dialogue design. 1

We present a variant of the WOz paradigm we refer to as incremental ablation. The new feature involves incrementally restricting the human wizard’s capacities in the direction of a dialog system. We lay out a data collection design with six conditions of user-system and user-wizard interactions that allows us to more precisely identify how to close the communication gap between humans and systems. We describe the application of the method to analysis of contexts in which ASR errors occur, giving us a means to investigate the problemsolving strategies humans would resort to if their communication channel were restricted to be more like the machine’s. We describe how we can use the methodology to collect data that is more relevant to a particular learning paradigm involving Markov Decision Processes (MDP).

Spoken dialogue systems (sdss)—computational systems that can engage in a dialogue with a human user about a restricted topic by understanding and reacting to spoken natural language—have a valuable potential not only as commercial systems that can handle useful tasks in real situations, but also as a test bed for semantic and pragmatic theories of dialogue interaction. Besides the difficulties inherent to any natural language understanding system (like e.g. ambiguity and context-dependence), sdss are faced with additional challenges that often derive from the lack of technical accuracy of their components, most prominently the automatic speech recogniser (ASR). For this reason, sdss are confronted with a great degree of uncertainty when processing user utterances. In those cases where understanding does not fail completely, a system will typically be able to form some hypotheses about the input received from the user. However, judging the quality of these hypotheses is itself a highly uncertain task, and finding answers to questions such as ‘Did the user really say X?’, ‘Did the user mean Y? ’ or ‘Is Z what the user intended me to do? ’ can be a very hard enterprise for a dialogue system. One of the crucial aspects that contributes to reducing uncertainty is the use of meaningful clarification and grounding strategies that are able to tackle the problems that the system encounters and, when appropriate, give feedback to the user about its internal representations.

This paper presents the first step in designing a speech-enabled robot that is capable of natural management of miscommunication. It describes the methods and results of two WOz studies, in which dyads of naïve participants interacted in a collaborative task. The first WOz study explored human miscommunication management. The second study investigated how shared visual space and monitoring shape the processes of feedback and communication in task-oriented interactions. The results provide insights for the development of human-inspired and robust natural language interfaces in robots. 1

Abstract. Due to the relatively simple structure of dialogues in previous spoken dialogue systems, discourse structure has seen limited applications in these systems. We investigate the utility of discourse structure for spoken dialogue systems in complex domains (e.g. tutoring). Two types of applications are being pursued: on the system side and on the user side. On the system side, we investigate if the discourse structure information is useful for various spoken dialogue system tasks: performance analysis, characterization of user affect and characterization of speech recognition problems. On the user side, we investigate whether the discourse structure information is useful for users of a spoken dialogue system through a graphical representation of the discourse structure.

To my parents, Paul and B.J., for encouraging both wonder and accomplishment. To my wife, Sarah, for her unflagging love, support, and understanding. To all my friends, who have helped in innumerable ways.

We study dependencies between discourse structure and speech recognition problems (SRP) in a corpus of speech-based computer tutoring dialogues. This analysis can inform us whether there are places in the discourse structure prone to more SRP. We automatically extract the discourse structure by taking advantage of how the tutoring information is encoded in our system. To quantify the discourse structure, we extract two features for each system turn: depth of the turn in the discourse structure and the type of transition from the previous turn to the current turn. The � 2 test is used to find significant dependencies. We find several interesting interactions which suggest that the discourse structure can play an important role in several dialogue related tasks: automatic detection of SRP and analyzing spoken dialogues systems with a large state space from limited amounts of available data. Index Terms: discourse structure, speech recognition analysis, spoken dialogue systems.