Language Access to Distributed Data with Error Recovery

Abstract

This paper discusses an effort in the application of artificial intelligence to the access of dat a from a large, distribut ed data base over a com puter network. A running system is described that provides real - time access over the ARPANET to a data base distributed over several ma c hines. The system accepts a rather wide range of natural l anguage questions a bout the data, plans a sequence of appropriate queries to the data base management system to ans wer the questi o n, determines on which machine(s) to carry out the qu erie s, establishes links to those machines over the ARPANET, monitors the pros ec ution of the queri es and recovers from ce rtain errors in execution, and prepares a relevant a nswer. In addition to the components that make up the demonstration system, more sophisticated functionally equivalent co mpon e nts are discussed and proposed.