A semantic framework to support corporate memory management in building construction

Corporate memory tools represent one way organizations can document, retrieve and utilize best practice and lessons learned in enhancing their performance. Using semantic systems in building these tools (along with database and/or Al-based systems) allows for more efficient representation of tacit knowledge. Such systems are based on a common ontology of the subject domain, where entities (such as actors, processes and products) are interlinked to represent the essence of the knowledge in the domain. The paper presents a taxonomy for building construction. The taxonomy includes 6000 concepts and was developed using OWL. It maps to existing classification systems to assure better coverage. The taxonomy is the first attempt, to present building construction knowledge in a semantic way. It also represents the foundations developing ontology-based corporate memory systems. To demonstrate the role and contribution of the proposed taxonomy a prototypical ontology for building construction was developed. Furthermore, a framework for agent-based system for supporting semi-automatic generation of reports such as lessons learned, work forms, and meeting agendas. Such agents allow organizations to capture and document its knowledge (in a taxonomy-complaint format) and to feed back post-project knowledge into new ones through access to lessons learned and through pre-defined meeting agendas. Access to these reports is done through semantic search according to the proposed taxonomy. Future research will develop a formal ontology and further develop the framework and implement it in actual organizations. (C) 2005 Elsevier B.V. All rights reserved.