A methodology for modeling an organization to determine and derive information systems requirements

This research deals with the development of an Organization Support System (OSS) whose goal is to improve the effectiveness of an organization. When one is interested in the content of an information system that supports an organization to achieve its objectives, then one is more concerned with the structure, characteristics, goal orientations and the strategies of the organization besides the basic problem of "how to manage well". As a consequence, for the development of an OSS, the analyst needs more sophisticated skills and a different array of tools to understand the integration of the organizational structures and processes. Furthermore, since the task is large the analyst has to work as a member of a team that includes users belonging to different parts of an organization. Each user has a different mental framework of the problems which differs from that of the analyst. This diversity of mental models further complicates the task. This investigation is based upon the belief that a successful analyst needs a support system for playing the roles of different users and for communicating the progress of analysis to his team members. The literature on requirement engineering suggests that a high percentage of design errors in computer based information systems can be attributed mainly to an imprecise statement of requirements. The most frequently observed shortcomings in the specification of system requirements are: (1) incorrect requirements; (2) inconsistent and incompatible requirements and; (3) unclear requirements. It is patently clear today that the above problems in requirement definitions arise due to the difficulty in understanding the supported organization and its environment. One confronts three major issues while studying an organization to determine requirements: (1) The understanding of an organization is difficult because of its inherent complexity and size. To tackle this complexity of size one must model the problem in such a way that one can study one aspect of the problem at a time while not losing the overall problem's context. This warrants a stepwise refinement approach to modeling. (2) The lack of formalism in describing the organization leads to ambiguity and inconsistency. The problem is compounded for large systems where a group of people work together to understand and determine requirements. A formal notation to describe and understand an organization and to determine requirements is needed to minimize the problems of communication, inconsistency and ambiguity. (3) In determining requirements for an OSS one cannot ignore organizational issues and policy implications. The software engineering approach of dealing with the issues of complexity and lack of formalism is not adequate. This research has focused on all of these issues. The outcome of this research is a methodology which consists of: (1) a set of guidelines in the form of a framework to study and characterize an organization for tackling the third issue described above, and (2) an automated tool which provides a formal notation to describe and build a conceptual model of an organization interactively. This tool helps in tackling the first and second issues described above. This methodology has been used to study the Bureau of Budget and Planning of the city of Atlanta. The use of the framework revealed organizational conflicts which would not have been brought out by any of the existing requirement definition methodologies. Despite the limited use of the methodology, our experience has clearly demonstrated the following: (1) The traditional way of collecting data by talking to people proved to be a fallacy. (2) There may be an inconsistency between the organizational structure and its policies. (3) A systematic top down study of the total organization avoids the design of a faulty system which may overload the organization functionally and lead to disfunctional behavior.