| Abstract: | I present a model for the capture and retrieval of organizational memory that can help overcome the information dissemination problem faced by network organizations. Network organizations are made of small, heterogeneous, product-oriented teams. In these organizations, the specialization of knowledge that has characterized functional hierarchies can be lost as the functional units in which this expertise was created and shared are dismantled. The preservation and dissemination of expert knowledge thus becomes a major issue to be addressed by these organizations, whose competitive advantage depends on this specialization. The problem of capturing project-related organizational memory is addressed with a 3 step solution: (a) the use of explicit work processes; (b) the development of models capable of representing these processes, the artifacts that are developed, and the rationale used during decision making; and (c) the implementation and deployment of computer tools that support the development processes selected and that, in doing so, record and document the project. I address the problem of accessing information from projects developed elsewhere in the organization and propose a solution based on monitoring the user's manipulation of information and creating hypotheses of his or her information needs. These hypotheses are then used as queries by a fuzzy information retrieval engine that produces a ranked list of documents that are suggested to the user as being relevant to his or her information needs. As the user manipulates these documents, the hypothesis is updated, triggering a new retrieval and the update of the list of suggested documents. The feasibility of the approach is demonstrated with the implementation of a prototype. |
