WIT is a tool designed specifically to automate the integration of Web-based, distributed databases in a manner that respects the desired autonomy of the database owners. Originally WIT was envisioned as a tool for supporting geographically distributed collaborative software development efforts. However, as it currently exists, the WIT prototype is general enough to support a wide range of distributed activities with a need for integrating distinct databases.
Web-based distributed databases are a style of Web database management systems that take advantage of both the distributed nature of the Web and the ease with which supplemental structure can be added to any Web-accessible data. By using a series of hyperlink-embedded HTML pages, these systems organize Web-accessible data into structured database content--all without alteration (or ownership) of that content.
The initial WIT prototype supports a distinctive type of Web-based distributed database structure that closely follows the classical semantic database model known as E/R, or the Entity/Relationship model. The novel idea of structuring Web-accessible data in this manner (i.e., superimposing E/R semantics) was conceived by researchers at Politecnico di Milano's CEFRIEL research center in Milan, Italy. This structure allows, among other things, the capability of explicitly expressing how pieces of database content (Web-accessible data) are related and linking them together for ease of access. Databases formatted with this structure can be integrated, or combined, into federated databases of the same structure. Further, the structure of the original databases, and their content, remains unaltered. Because the resulting database uses the same organizing strucure as the individual databases, these integrated databases can be further integrated, using WIT, or manipulated by the database systems that rely on that structure, such as the Labyrinth DBMS developed by researchers at Politecnico di Milano's CEFRIEL research center in Milan, Italy.
A more advanced prototype is currently under construction that focuses on increasing the semantic-related analytical capabilities of the tool. This capability will further automate the process of confirming that databases participating in a federation are aware of each others' terminology when describing specific kinds of entities or the relationships among them. It will also enable users to be more selective about the specific information to be integrated by the WIT system.
This work is sponsored by the Defense Advanced Research Projects Agency and the Air Force Material Command, Rome Laboratory, under Contract Numbers F30602-94-C-0253 and F30602-98-2-0163.