Integrated queries to existing bibliographic and structured databases

Abstract

It is widely accepted that future digital library applications have to be built upon different kinds of database servers to draw upon different forms of data, including bibliographic, text, multimedia, and structured data. In this paper, the problem of integrating existing public bibliographic databases and structured databases which reside at different locations in the network is addressed. Although bibliographic data is semistructured, its attribute set is often determined by an international standard known as MARC. To unify bibliographic and structured data, the well-known SQL was extended to model bibliographic related attributes and queries. In particular, a new data type was added to model attributes in the bibliographic database. Specialized functions have also been designed that can be used to formulate queries on bibliographic data, as well as queries on both bibliographic and structured data. By combining the SQL extensions with other query language extensions that access general text data, it is believed that a versatile query language layer can be made available for future digital library application development.