BUSTER - An Information Broker for the Semantic Web

Among the most fundamental requirements for Semantic Web applications is their ability to access and to interpret information sources that are geographically distributed, as well as highly heterogeneous with respect to syntax, structure and semantics (Visser and Klein 2002). In order to be able to decide whether an information source is relevant to a specific information request and to resolve semantic heterogeneities during the subsequent integration of this information source, Semantic Web applications must have access to metadata that describe the semantics of an information source. The Bremen University Semantic Translator for Enhanced Retrieval (BUSTER) is an information broker middleware that was built to tackle the challenges of the Semantic Web. BUSTER is able to access heterogeneous and distributed information sources and to assess their conceptual, spatial, and temporal relevance with respect to a specific information request. BUSTER can also be used to integrate heterogeneous information through the resolution of structural, syntactical, and semantic heterogeneities. In this article, we focus on the application of BUSTER as a tool for intelligent search and information retrieval in the Semantic Web. The BUSTER search module supports the specification of queries of the type concept @ location in time. In addition to the conceptual semantics, the system evaluates the spatial as well as the temporal relevance of an information source. A typical example for an information request would be: “Find a conference-hotel near the Thüringer Wald for the period between Christmas and New Year 2003”. In order to be able to reason about conceptual (“Which concepts are subsumed by the concept specified in the query?”), spatial (“Which regions are in or near the specified geographic region?”), and temporal relevance (“Which dates fall within the given time period?”), BUSTER utilizes metadata that provide formal descriptions of the respective context of an information source. While well-established methods to formalize and to reason about conceptual semantics could be used, the inclusion of spatial and temporal aspects demanded the development of new representation schemes and reasoning mechanisms. Given the distributed structure of the Semantic Web and its implicit open world assumption, a major challenge was to provide a framework for the interoperability and integration of formal conceptual, spatial, and temporal representations. In the following, we will give a brief introduction to the methods and solutions we developed to meet these challenges. BUSTER – An Information Broker for the Semantic Web