Multiple language capability in an interactive system.

摘要

Announcements from a voice messaging system that either prompt a user for action or provide information to a user are constructed independent of the language in which the announcement is provided to the user. In particular, each announcement is treated as a "semantic expression" having an announcement identifier (310). This announcement identifier is mapped into a plurality of rules (315), each rule providing instructions on how to construct the announcement for a particular language (320). The rules can recursively call other rules to provide additional flexibility in constructing a particular announcement. <IMAGE>