Towards A Meta-Model for Representing Organizational Memory Balasubramaniam Ramesh Naval Postgraduate School, Monterey, CA 93943 ramesh@nps.navy.mil Abstract For each of the stakeholders involved in complex organizational processes, some useful support can be provided by recording in some structured fashion relevant portions of organizational memory. For this purpose, components of OM should include rich contextual information not only to provide detailed viewpoints of individual decision makers, but also to identify interdependencies among differing perspectives and viewpoints. An effective scheme for the capture and use of OM includes both formal and informal information. We describe a meta-model that can be used to represent comprehensive OM schemes. We describe the capabilities of a prototype OMS that provides a variety of facilitates for the capture, maintenance and use of OM. 1. Introduction Organizational memory provides the means by which organizational knowledge from the past can influence the present organizational activities [36]. It is “stored information from an organization’s history that can be brought to bear at present decisions” [40]. Recent literature emphasizes the importance of mechanisms to capture and manage rationale in situations involving large number of participants involved in complex organizational processes like large scale systems development [8]. In collaborative activities involving a large number of participants, each having a different set of goals and priorities, maintaining a comprehensive history can be invaluable. Such a history is an important component of OM and includes various types of memory which Stein and Zwass [36] call Group/team memory, design rationale/discussion memory, and project memory. In this paper, we take this limited, task specific view of Organizational Memory, recognizing that a comprehensive organizational memory includes information retained by individuals, culture, transformations, structures and ecology [40][36]. A variety of stakeholders involved in organizational decision processes bring together their often unique viewpoints and expertise. These range from top management providing inputs on the nature of the problem and its impact on the organizational goals, to middle managers who help articulate objectives and constraints, to lower level workers that implement them. In contexts where tightly integrated decision processes across functional boundaries cannot be easily developed and implemented, organizational memory that provides a “loose coupling” between relevant decision situations can greatly enhance shared understanding across these boundaries. Coordination among the various stakeholders (design, marketing, finance, manufacturing etc.) which is critical to success can be facilitated with organizational memory that integrates their various perspectives. A basic premise in our work is that for each of the stakeholders involved in such processes, some useful support can be provided by recording in some structured fashion relevant portions of organizational memory. Recent studies confirm that capturing organizational memory is especially important in large-complex activities for the following reasons: The context in which key decisions are made will be lost when different teams are engaged in various aspects of the organizational process. Critical errors are commonly made in formulation and resolution of decisions, but they are often unnoticed in the absence of comprehensive organizational memory [37] In the absence of reliable organizational memory, work groups engage in repetitive discussion and resolution of the same issues [43]. When stakeholders with differing expertise, perspective and viewpoints are involved, key decisions are often misunderstood and misinterpreted [42]. Radermacher [23] argues that complex decision making situations involve a constantly changing environment and preferences. Further, these situations are also characterized by incomplete and even inaccurate information. In such complex decision situations, a comprehensive organizational memory of the context in which decisions are arrived at and the ability to understand and analyze its implications when it changes will be extremely helpful The need to capture and reason with the information regarding the “context” is even more pronounced when the decision making includes participants from various “functional” domains. OM that provides the ability to modify and manipulate decisions in response to changing requirements and assumptions are valuable in diagnostic problem solving [39]. Silverman [32][33] supports this view, emphasizing the importance of facilities to simulate, 1060-3425/97 $10.00 (c) 1997 IEEE Proceedings of The Thirtieth Annual Hawwaii International Conference on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE