A grid-based architecture for personalized federation of digital libraries

Federated Digital Libraries (FDLs) unify metadata from geographically distributed digital libraries and implement advanced information retrieval services for the users. Currently, the growing number of users and digital libraries on the World Wide Web (WWW) as well as the use of computationally intensive indexing and search algorithms causes scalability and performance problems. In order to build efficient and effective personalized FDLs, we developed a novel architecture and a system prototype, which is based on state-of-the-art grid technology. When using the system, the user selects a set of geographically distributed DLs. Then, the system harvests metadata from the selected DLs and computes and stores an index. Following this procedure, the user can use the generated index for searching her or his personalized FDL. It is shown that the proposed architecture is generic and scalable, and it can, as such, accommodate a number of different FDL applications.