Information systems with partially unknown purposes

In “traditional” information systems design theory it is usually assumed that the information system to be designed has, at least in principle, a well-defined purpose, that is, a purpose from which it is possible to derive, among other things, the information needs to be satisfied by the information system. However, there is a category of information systems of growing practical significance, where the purposes are partially unknown. Some examples of types of information systems which typically belong to this category are decision support systems, executive information systems, statistical information systems, scientific information systems, and metainformation systems. When an information system belonging to this category is going to be designed, it is usually impossible to specify its usage with great precision. The business decisions, research problems, or whatever kind of activity it is that the information system is going to support may not yet even have been thought of; they may lie a long time ahead, and by then many things may have changed within the area of interest and its environment, as well as in the focus of interest of the information system users. This paper discusses some of the methodological challenges posed by information systems with partially unknown purposes.