Knowledge-Intensive Query Processing

Innovative query interfaces to knowledge and database systems must go beyond simply returning the requested information. They must be capable of producing intentional answers when a description improves the understanding of an answer [Mot94], producing conditional answers when no one answer matches the conditions of a query, and using ontological information in processing a query. They should be able to call upon stand-alone reasoning modules that are most suitable for a given query. When answering a question involves reasoning beyond a simple lookup, the system must be able to explain the answer to the user. We are building a question answering system with these objectives. The heart of the system is a knowledge base (KB) and a collection of reasoning methods. The KB is being constructed by a combination of manual and semiautomatic methods. The reasoning methods include conventional database query processing, frame-based reasoning, and full first-order theorem proving. The performance of this system will be tested on the Crisis Management Benchmark (CMB), which defines a collection of queries of interest to a crisis analyst. We begin the paper by a description of the CMB. We describe the architecture of our system and then sketch some design ideas for two of its components.