Integrating Distributed Data Sources Using Federated Objects

We address the problem of integrating heterogeneous data sources for collaborative environments such as Groupware systems. Pre-existing related distributed data is modeled by a composite object, where each data source is treated as a fragment of the composite object. We define a collection of methods which allow manipulation of the related distributed data in a controlled and consistent manner. The composite object created in this manner is referred to as federated object, and it provides a pragmatic approach to data integration. Our approach is different from the approach of integrating heterogeneous database systems, as we do not require each data source to have full database system capabilities. Instead, for the collaborative application environments the data sources can be simple files, application gen~ eraLed binary files, or commercial data servers. The federated objects are especially suitable for computer-supported-cooperative work (CSCW), where coordination and communication among a group of people is required, and the type of collaboration is dynamic in nature. In this paper, we discuss the general properties of federated objects, present implementation principles for the required mappings, and show how federated objects can be used to implement a simple report writer application that needs to access several distributed data sources. A fully developed federated object mechanism relies on a "toolkit" approach, where federated objects are constructed from existing data sources using a C++ class library, which provides the mechanisms needed for access and manipulation of a wide variety of data sources.