Creating domain specific metadata for scientific data and knowledge bases

The authors examine a particular type of domain-based metadata for which a complete ready-made conceptual framework is not and probably cannot be directly supported in predefined data models of knowledge-based representation paradigms. The data or knowledge base designer thus has the burden of sometimes working outside a familiar and otherwise appropriate model. The authors discuss issues arising in formulating certain domain-based metadata extensions and provide guidelines for developing them. These results derive from a continuing effort to create a knowledge base to support research on biological organisms. The added semantics are used to support a more flexible and reliable system for identification