A context resolution server for the galaxy conversational systems

The context resolution (CR) component of a conversational dialogue system is responsible for interpreting a user’s utterance in the context of previously spoken user utterances, spatial and temporal context, inference, and shared world knowledge. This paper describes a new and independent CR server for the GALAXY conversational system framework. Among the functionality provided by the CR server is the inheritance and masking of historical information, pragmatic verification, as well as reference and ellipsis resolution. The new server additionally features a process that attempts to reconstruct the intention of the user given a robust parse of an utterance. Design issues are described, followed by a description of each function in the context resolution process along with examples. The effectiveness of the CR server in various domains attests to its success as a module for context resolution.