A Metadata Integration Framework for

Semantic heterogeneity is the problem of determining whether a set of documents describe the same information. Two different documents may belong to different document types, using different document schemas, which makes it difficult to determine whether the two documents describe the same kinds of information. To deal with the problem, metadata can be used to provide a uniform interface to a corpus of documents with heterogeneous document types, analogous to the integration of autonomous databases in a federated database arrangement. This service is provided through an integration architecture, in which metadata conforming to a unified metadata schema are used to describe the information in documents common among all document types. With the integration architecture in place, advanced query mechanisms can be built on top of it to allow integrated, transparent queries to be formulated. This thesis demonstrates how metadata can be used to solve the semantic heterogeneity problem. Insights are presented that describes the kinds of metadata that are suitable for capturing data common among the variety of document types encountered in the GIS community, with a description for a model of an integration architecture to best capture the common metadata, along with an appropriate metadata classification to meaningfully structure the metadata. Information is presented on how the extracted metadata can be subsequently used in the creation of concepts in an ontology-like structure. Finally, a basic architecture establishes how to per-