Implementation of interoperability in large multidatabases

Sharing information among databases has traditionally been at the data level. In a large network of databases and with the current technology, direct data sharing, i.e. sharing actual data in one single step, is no longer tractable or even possible. Another layer of data sharing has to be supplied before direct data sharing can take place. The size issue is a major factor in complicating issues of autonomy and heterogeneity. In this new environment, users need to be incrementally and dynamically informed about available information and where it is located. A static approach to user education cannot be reasonable in such a large environment. This approach relies on building a new layer that is responsible for organizing databases in suitable and flexible conglomerations. This layer is implemented using special-purpose databases. These are called co-databases.<>