What ' s not in a name : Some Properties of a Purely StructuralApproach

One approach to integrating knowledge bases is based on nding assertions that relate the expressions in the constituent terminologies. For knowledge bases with many terms this task requires computer support. We set up a formal framework for merging Description Logic TBoxes, and then explore the limits of a purely structural approach to the problem of nding interrelationships between knowledge bases. Some theoretical notions are empirically examined in a real medical ontology (galen). Suppose we are given two terminologies (also called ontologies, conceptual models) | collections of terms and their meanings as used in some universe of discourse (UofD). Our task is to integrate/align them into a new terminology that captures the meaning of the original terms and their interrelationships. This problem is motivated by several application scenarios. First, terminologies have been developed independently by multiple research groups in a variety of application areas, including medicine and natural language processing; for example, quite recently, several ontologies about logistics were used as part of the HPKB project. In each of these cases there have been eeorts at merging these terminologies 16, 20, 17]. Second, there are many examples of heterogeneous information sources, such as federated databases or semi-structured data on the internet, which keep information about aspects of some single domain of discourse (e.g., movies), and