Knowledge has become a key to success for a business in the era of knowledge economy. For a business that relies on its intellectual capital as the weapon to compete in its market, a learning organization should be developed to manage its knowledge in an effectively way. In the learning organization, the elicitation, categorization, distribution, and sharing of knowledge are the major requirements. In this project, we study the computationally plausible ways of supporting the requirements. We attempt to bridge the gaps between current information technology and knowledge management. Based on our previous work in knowledge management, information retrieval, and multiagent systems, we plan to develop an agent-based model CABKM to collaboratively elicit, categorize, and distribute knowledge, which is represented in document forms and shared among employees in a business. To link employees with knowledge pieces, an agent is designed to correspond to a knowledge category and a department of a business. Thus each agent is responsible to store knowledge pieces (for its corresponding category) and distribute knowledge pieces (for its corresponding department of employees). In practice, knowledge categories and departments are often organized in a hierarchical way. We thus explore how the hierarchical structure of the knowledge categories may be employed to construct the agents automatically. Given a knowledge piece or a request, the agents work together to identify its categories by considering the contexts of the knowledge piece. Accordingly, subsequent knowledge management activities may be autonomously conducted by the activated agents. Preliminary studies and experiments show that the feasibility of CABKM may be expected. In this project, we will theoretically and empirically evaluate the performance and contributions of CABKM.
Keywords: Knowledge Management, Collaborative Agents, Knowledge Categorization, Knowledge Elicitation, Knowledge Sharing
Back to Research Project