Problems With Domain-Independent Natural Language Database Access Systems

In the past decade, a number of natural language database access systems have been constructed (e.g. Hendrix 1976; Waltz et al. 1976; Sacerdoti 1978; Harris 1979; Lehnert and Shwartz 1982; Shwartz 1982). The level of performance achieved by natural language database access systems varies considerably, with the more robust systems operating within a narrow domain (i.e., content area) and relying heavily on domain-specific knowledge to guide the language understanding process. Transporting a system constructed for one domain into a new domain is extremely resource-intensive because a new set of domain-specific knowledge must be encoded.In order to reduce the cost of transportation, a great deal of current research has focussed on building natural language access systems that are domain-independent. More specifically, these systems attempt to use syntactic knowledge in conjunction with knowledge about the structure of the database as a substitute for conceptual knowledge regarding the database content area. In this paper I examine the issue of whether or not it is possible to build a natural language database access system that achieves an acceptable level of performance without including domain-specific conceptual knowledge.