Knowledge Management in Scientific Domains

Scientific domains are characterized by substantial amounts of complex data, many unknowns, lack of complete theories, and rapid evolution. In many areas of decision making, much of the reasoning process is based on experience rather than on general knowledge. Experts remember positive cases for possible reuse of solutions, but negative cases are also useful for avoiding potentially unsuccessful results. Thus, storing and reasoning with experiences facilitates efficient and effective knowledge management. In general, knowledge management systems support representation, organization, acquisition, creation, usage, and evolution of k_nowledge in its many forms. Complex scientific domains require: (1) multimodal representations that support application-domain richness, expressibility and domain-knowledge evolution, (2) effective organization of l~owledge for efficient access to information, and (3) decision-support and analysis tools. We show how a case-based reasoning system can be used for knowledge management in structural biology. Namely, we describe a multimodal and multimedia system for managing crystallization experiences.