LUKE: An Experiment in the Early Integration of Natural Language Processing

LUKE: AN EXPERIMENT IN THE EARLY INTEGRATION OF NATURAL LANGUAGE PROCESSING David A. Wroblewski, Elaine A. Rich MCC Human Interface Laboratory 3500 West Balcones Center Drive Austin, Texas 78759 Abstract Luke is a knowledge editor designed to support two tasks; the first is editing the classes and relations in a knowledge base. The second is editing and maintaining the semantic mapping knowledge neccesery to allow a natural language interface to understand sentences with respect to that knowledge base. In order to emphasize design decisions shared between the two tasks, Luke provides facilities to concurrently debug the application and the natural language interface. Luke also makes natural language available in its own user interface. This makes it possible for a knowledge base builder to exploit natural language both as a way of locating desired concepts within the knowledge base and as a a way of doing consistency checking on the knowledge base as it is being built. base, Luke makes natural language available in its own interface. This makes it possible for the knowledge base builder to exploit natural language both as a way of referring to objects in the knowledge base and as a way of doing consistency checking on the objects themselves. In this paper, we will describe both what Luke does and how doing that supports this productive view of the interaction between building a knowledge based system and building an associated natural language interface. Background And Motivation Introduction Luke is a knowledge base editor that has been enhanced to support entering and maintaining the semantic mappings needed by a natural language interface to a knowledge base. Thus Luke supports a team of system builders who are simultaneously building a knowledge- based program and building a natural language interface to that program. It makes sense for a single tool to support both of these efforts because the efforts themselves are logically intertwined in two important ways, both of which result from the fact that the application program and its NL interface must share a single knowledge base. (This sharing is necessary because otherwise the NL system will not be able to communicate with the application). The first way in which the two efforts Luke supports are related is that, although they produce two systems that are different and may thus place different demands on their associated knowledge bases, both must share a single such knowledge base. By supporting the early integration of the application program and the NL interface as this single knowledge base is being built, Luke helps to ensure that it will be adequate, with respect to both its content and its structure, to support both these target tasks. The second way in which the two system building tasks are related is that one can support the other. By associating natural language with concepts as they are entered into a knowledge A Model Of Semantic Analysis All of the following discussion is based on a model of semantic analysis similar to that proposed in (Hobbs, 1985). …

Content truncated to 500 words