DeepListener: Harnessing Expected Utility to Guide Clarification Dialog in Spoken Language Systems
- Eric Horvitz ,
- Tim Paek
6th International Conference on Spoken Language Processing (ICSLP 2000), Beijing |
We describe research on endowing spoken language systems with the ability to consider the cost of misrecognition, and using that knowledge to guide clarification dialog about a user’s intentions. Our approach relies on coupling utility-directed policies for dialog with the ongoing Bayesian fusion of evidence obtained from multiple utterances recognized during an interaction. After describing the methodology, we review the operation of a prototype system called DeepListener. DeepListener considers evidence gathered about utterances over time to make decisions about the optimal dialog strategy or realworld action to take given uncertainties about a user’s intentions and the costs and benefits of different outcomes.
Information Agents: Directions and Futures (2001)
In this internal Microsoft video, produced in 2001 and released publicly in 2020, research scientist Eric Horvitz provides glimpses of a set of research systems developed within Microsoft’s research division between 1998 and 2001. Projects featured in the video include Priorities, Lookout, Notification Platform, DeepListener, and Bestcom. The projects show early uses of machine learning, perception, and reasoning aimed at supporting people in daily tasks and at making progress on longer-term missions of augmenting human intellect. The efforts are thematically related in their pursuit of broader understandings of people and context, including a person’s attention, goals, activities, and location, via multimodal signals, involving the analysis of multiple streams of information. Several of the prototype systems were built within the Attentional User Interface (AUI) project, which…