The XCALIBUR Project, A Natural Language Interface to Expert Systems and Data Bases
Any type of content formally published in an academic journal, usually following a peer-review process.
The inevitable proliferation of expert systems underscores the need for robust, friendly interfaces requiring minimal user training. The objective of the XCALIBUR project is to meet this need by providing natural comprehension and generation in the context of a focused mixed-initiative dialog. The XCALIBUR architecture is discussed, including its three central components (parser, generator and information manager), its methods of handling ellipsis and imperfect input, and its relation to the underlying expert system.