Organizing the concepts of science: Science ontologies and the semantic Web

An ontology is an explicit, formal representation of the concepts, objects or other entities in a particular domain and the relationships among them. These advanced knowledge organization structures are increasing in number and relevance as computer to computer processing of information becomes more prevalent. They have been developed in many scientific disciplines where computer agents are being developed to support the scale of work, including bioinformatics, genetics, earth and space science, medical research, and engineering. Ontologies have been identified as a key technology for the advancement of the semantic web, which seeks to provide more precise and relevant information for human and computer use. The session focuses on ontologies in three science areas: aerospace and space science, cancer research, and agriculture. These case studies show how research is meeting practice. For each case study domain, sample ontologies and how they differ from other knowledge organization structures such as thesauri or topic maps are presented. Examples show the types of relationships expressed and how these relationships are manifested in real ontologies. The specific decisions made by a particular community regarding the tools selected for creating and managing the ontology (both open source and commercial), the interrelationship of ontologies with other knowledge organization systems in that domain, methods for gathering concepts for the ontology from members of the community, and the use of the ontology to support searching in large structured and unstructured environments such as the Web are described. The usefulness of ontologies in these domains has resulted in the development of numerous specific ontologies in each domain. Efforts are underway to identify, describe, manage, access and integrate these various ontologies in support of semantic web, Grid computing, and knowledge sharing efforts. The major issues related to interoperating and integrating very specific ontologies to bridge domains and projects within an enterprise are identified. The research into technologies and standards to support ontology integration is discussed. The eXtended Metadata Registries (XMDR) Project, sponsored by the Department of Defense, Lawrence Berkeley Laboratory, and the Environmental Protection Agency, and other approaches to integration are highlighted.