Case Study 1: Webfindit Prototype

Advances in wide area networking are making the goal of achieving a worldwide interoperable database system closer than ever. In a large network of interoperable autonomous heterogeneous databases, the architecture should be flexible enough to allow negotiation to take place to establish grouping of databases. In such an environment, it is important that databases be made aware of other participating databases in an incremental fashion. Furthermore, users should be educated about the space of information in an incremental and dynamic fashion. These goals are to be achieved under the major assumptions that participating databases keep their autonomy intact as much as possible and that the overhead in bridging heterogeneity be proportional to the level of sharing. In this chapter, we present a system for describing, locating, and accessing data in large networks of databases. The use of flexible constructs such as service links, coalitions, and co-databases facilitates data organization, discovery, and sharing. A language supporting the definition and manipulation of these constructs is also proposed. We propose an implementation using CORBA (Common Object Request Broker) and Web technologies to support scalable and portable architecture interconnecting large networks of heterogeneous and autonomous databases.