Semantic assistance nets

Improving the semantic level of the information handling provided by personal computers is likely to be a way both to use the increasing hardware capacities of these machines and to improve their economic impact. While this goal is widely recognized as abstractly desirable, it has not been clear how to accomplish it. This paper presents a design for organizing resources that are presently available, or that soon will be, to obtain systems that should provide significantly improved aids to understanding in fields where there is a substantial base of empirically verified information that involves nonnumeric relationships. Laboratory sciences appear to provide the fields with the greatest scope for such development at present, but there may also be opportunities to use such systems in investment and entrepreneurship. The resources to be assembled include frame-oriented databases and their associated ontological engineering organizations, high-bandwidth digital communications networks, microprocessor systems for vectorized numerical processing and object-oriented processing, larger units for these same purposes, high-capacity storage media, and network management software to coordinate the use of special hardware resources. The role of the personal computer in such a system will be as a combination of a vehicle for navigation through the information structures thus provided and as a node where information received over networks will be integrated with locally produced data. There have been many previous papers about topics related to those in this paper, but the focus of the present discussion is on integrating diverse information to outline the characteristics of the new systems, the prospect of successful development based on sound mathematical principles, and the specificity of the concept for important new fields of application.