A maximum entropy approach to natural language processing
Computational Linguistics
Cross-lingual relevance models
SIGIR '02 Proceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval
Towards developing general models of usability with PARADISE
Natural Language Engineering
AI Magazine - Special issue on achieving human-level AI through integrated systems and research
Human Computer Interaction in Virtual Standardized Patient Systems
Proceedings of the 13th International Conference on Human-Computer Interaction. Part IV: Interacting in Various Application Domains
A Virtual Tour Guide for Virtual Worlds
IVA '09 Proceedings of the 9th International Conference on Intelligent Virtual Agents
Building effective question answering characters
SigDIAL '06 Proceedings of the 7th SIGdial Workshop on Discourse and Dialogue
Ada and grace: toward realistic and engaging virtual museum guides
IVA'10 Proceedings of the 10th international conference on Intelligent virtual agents
Toward learning and evaluation of dialogue policies with text examples
SIGDIAL '11 Proceedings of the SIGDIAL 2011 Conference
| Hi-index | 0.00 |
This paper presents an analysis of how the level of performance achievable by an NLU module can affect the optimal modular design of a dialogue system. We present an evaluation that shows how NLU accuracy levels impact the overall performance of a system that includes an NLU module and a rule-based dialogue policy. We contrast these performance levels with the performance of a direct classification design that omits a separate NLU module. We conclude with a discussion of the potential for a hybrid architecture incorporating the strengths of both approaches.