With the increased deployment of conversational systems, new technical challenges and limitations appear. Client devices do not have the processing power and memory requirements to perform the necessary recognition and presentation tasks. Also, networks do not have enough bandwidth to rapidly exchange the data files needed by the conversational engines. Assuming that these two problems are solved, vendors and service providers often do not wish to exchange such data files which may be considered as intellectual, business logic and technology crown jewels. Distributed architectures solve these issues, if implemented in appropriately managed networks to guarantee quality of service for each active dialog. In this paper, we introduce some aspects of conversational networking: distributed speech recognition (DSR), distributed conversational architecture and conversational protocols for transport, coding and control. These are building blocks of a conversational networking solutions and distributed multi-modal browsers.
Cite as: Maes, S.H., Chazan, D., Cohen, G., Hoory, R. (2000) Conversational networking: conversational protocols for transport, coding, and control. Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000), vol. 2, 198-201, doi: 10.21437/ICSLP.2000-243
@inproceedings{maes00b_icslp,
author={Stéphane H. Maes and Dan Chazan and Gilad Cohen and Ron Hoory},
title={{Conversational networking: conversational protocols for transport, coding, and control}},
year=2000,
booktitle={Proc. 6th International Conference on Spoken Language Processing (ICSLP 2000)},
pages={vol. 2, 198-201},
doi={10.21437/ICSLP.2000-243}
}