With the current stampede toward electronic commerce, businesses no longer have the luxury of setting up trading-partner-specific communication systems. Businesses want to communicate and do business with people and companies with whom they have not had any prior contact. They also want to build more extensive linkages with their best trading partners and in their outsourcing relationships. Companies have built these links using electronic data interchange (EDI). EDI cannot be implemented easily, quickly, or inexpensively, and is therefore inappropriate for many companies and people to adopt. An application based on the technologies proposed in this paper would have, already built in, a certain capability for understanding messages plus a flexibility that would allow the system to grow with few restrictions provided by the underlying communication system. The benefit is that almost any message that is structured correctly can be handled to some extent further, more complex message handling for new messages can be defined incrementally.
In this paper we discuss a new kind of information system that helps people be ready for information work and locate documents. This system differs from a traditional information retrieval system by relying extensively on descriptions of both how a document is used and the purposes it is used for. These descriptions are gathered as the document is electronically used and manipulated (e.g., by a word processor or e-mail system). A formal language represents this information.
This paper motivates the need for system-level message management software. It begins by considering information flows in the workplace as a source of potential gains in efficiency. We next investigate work-flow automation and electronic data interchange (EDI) as indicative of current technologies applied to work processes and message management. Having described current technology and our vision of work processes, we propose an alternative, general-purpose, software technology for supporting application-to-application communication. Problems of EDI, of process-to-process communication, and of describing information items are discussed in terms of the communication problems they present. We then justify the need for this kind of software and lay out the criteria (or plausibility conditions) for evaluating a proposal for this sort of system software. The use of a formal communication language is proposed as a common solution to these problems. This proposal is examined in the context of the EDI problem, in order to demonstrate how the proposal might work in practice. Practical benefits of the proposal are discussed that highlight the impact such a technology might have on business practices. The proposed solution is measured against the plausibility conditions presented earlier in the paper; it is found to be sufficient in some cases and in need of further investigation in others. We then discuss the industrial-organizational implications of the availability of such a technology, and hypothesize that it would affect the number and form of cooperative business relationships as well as their scope and depth. We also hypothesize that it would provide advantages to those firms that quickly adopt the technology.