Abstract:With the fast growth of image database, managing image contents is becoming an increasingly important part of the development of organizational memory information systems. However, effective and precise image retrieval still remains a problem because of the extreme difficulty in image understanding, and high-level knowledge required. In this paper, a new semantic image retrieval method using the analytic hierarchical process is proposed, to provide a new way to incorporate experts' judgment into the system and to evaluate the fitness of a semantic description used to represent an image. The main aim of the proposed method is to facilitate mapping the perceptions of low-level features of the knowledge users to higher-level organizational knowledge. A prototype of this system is implemented and experimental results show that the performance of the proposed method is excellent when compared to traditional techniques. The implications, limitations, and future works of the proposed method for the development of organizational memory information systems are discussed.