Language Access to Distributed Data with Error Recovery

This paper discusses an e f f o r t in the a p p l i c a t i o n of a r t i f i c i a l i n t e l l i g e n c e to the access of data from a l a rge , d i s t r i b u t e d data base over a computer network. A running system is described that provides rea l t ime access over the ARPANET to a data base d i s t r i b u t e d over several machines. The system accepts a ra ther wide range of na tu ra l language questions about the data, plans a sequence of appropr ia te queries to the data base management system to answer the quest ion , determines on which machine(s) to car ry out the quer ies , es tab l ishes l i n k s to those machines over the ARPANET, monitors the prosecut ion of the quer ies and recovers from ce r ta in e r ro rs in execut ion, and prepares a re levant answer. In add i t i on to the components tha t make up the demonstration system, more soph is t i ca ted f u n c t i o n a l l y equiva lent components are discussed and proposed.