COLING '02 Proceedings of the 19th international conference on Computational linguistics - Volume 2
Processing broadcast audio for information access
ACL '01 Proceedings of the 39th Annual Meeting on Association for Computational Linguistics
| Hi-index | 0.00 |
Intelligent communication often requires context information to trigger proper communication services. In many cases, names, locations, as well as activities and status of communication participants are used to enable context-aware communication. In this paper, we propose a new context-aware communication paradigm, namely content-aware communication, which infers context information based on the content of ongoing conversations. New communication services can then be introduced by utilizing the content and the inferred information. Content-aware communication employs Automatic Speech Recognition (ASR) to acquire conversation content, Information Extraction (IE) to help identify useful context information, and Information Retrieval (IR) to find related information. The existing ASR and IR technologies can already provide applicable approaches to enable content-aware communication for a single user on his or her personal computer. However, there are still very few existing content-aware voice communication services and also lacks a secure and scalable way to integrate different technologies and resources for enterprise wide deployment of the services. In this paper, we first categorize enterprise content-aware communication services and illustrate some new content-aware services. We then define an architecture with distributed media processing and centralized call control to manage enterprise content-aware communications. This architecture also helps manage feature interactions when integrating content-aware services with other enterprise communication features. In addition, we allow enterprise users to experience content-aware communication on different devices and in different modalities.