A computer architecture to support natural full text information retrieval

An information retrieval system based on full text handling, coupled with a test surrogate front-end, is described. An architecture for implementing a 100-billion character textual database in a multiuser environment, with capability to conduct string-searching at an effective rate of 200 million characters/s is described. The search rate is independent of the number and complexity of the queries (up to the limiting size of the system), and the query language provides every known string search feature. The heart of this system is the fast data finder (FDF). The proposed architecture consists of off-the-shelf hardware and a small to moderate amount of software development. It is a completely general-purpose text handling system, independent of application or data