Detecting, Repairing, and Preventing Human-Machine Miscommunication

Article excerpt

* This article summarizes a workshop entitled "Detecting, Repairing, and Preventing Human-Machine Miscommunication," held on 4 August 1996 in Portland, Oregon. The author presents the significant issues raised during the four specific workshop sessions.

Any system that communicates must be able to cope with the possibility of miscommunication -- including misunderstanding, nonunderstanding, and misinterpretation. Research related to achieving robust interaction is an important subarea in AI. Early work concerned the correction of spelling or grammatical errors in a user's utterance so that the system could more easily match them against a fixed linguistic model; work has also been done in the area of speech recognition, attempting to find the best fit of a sound signal to legal sequences of linguistic objects. All these approaches have assumed that the system's model is always correct. More recently, researchers have been looking at detecting and correcting errors in the system's model of an interaction. This work includes research on speech repairs; miscommunication; misunderstanding; nonunderstanding; and related work in planning, such as plan misrecognition and plan repair.

The Workshop on Detecting, Repairing, and Preventing Human-Machine Miscommunication brought together researchers interested in developing theoretical models of robust interaction or designing robust systems. We were particularly interested in results drawn from experiments and applications that use speech as their primary modality of interaction. Some experiments involving multiple modalities were also discussed.

The workshop was organized into four sessions: (1) "Empirical Data Regarding the Occurrence of Miscommunication," (2) "Strategies for Identifying Potential Causes of Break- downs," (3) "Knowledge Representation and Reasoning about Miscommunication," and (4) "Repair in Spoken Language Systems." These sessions represent a progression from work that clarifies the problem of miscommunication to work that describes the strategies used to repair miscommunication. I review the most significant issues raised by the participants at these sessions.

In the session on empirical issues, the participants discussed various approaches to empirically evaluating hypotheses about human-machihe miscommunication. The approaches differed in two dimensions: First, experimenters specified different environments of the interaction by selecting different modalities of interaction or distributions of initiative (control) for the interaction. Second, experimenters selected a method for eliciting data; they used the computer to mediate between two humans, participate in the collaborative performance of some task with the user, or simulate an error-prone user interface to an implication. …


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.