The use of belief networks for mixed-initiative dialog modeling

This paper describes the use of Belief Networks for mixedinitiative dialog modeling within the context of the CU FOREX system [1]. CU FOREX is a bilingual hotline for real-time foreign exchange inquiries. Presently, it supports two separate interaction modalities: a direct dialog (DD) interaction, which is systeminitiated for novice users; as well as natural language (NLS) shortcuts, which is user-initiated for expert users. In this work, we propose to use Belief Networks (BNs) to automatically govern the model transitions for mixed-initiative interactions. Furthermore, we hope that our approach can reduce the amount of handcrafting involved in the development of current mixedinitiative dialog models, to ease portability across different applications domains.