Automated Agent Ontology Creation for Distributed Databases

摘要

In distributed database environments, the combination of resources from multiple sources requiring different interfaces is a universal problem. The current solution requires an expert to generate an ontology, or mapping, which contains all interconnections between the various fields in the databases. This research proposes the application of software agents in automating the ontology creation for distributed database environments with minimal communication. The automatic creation of a domain ontology alleviates the need for experts to manually map one database to other databases in the environment. Using several combined comparison methods, these agents communicate and negotiate similarities between information sources and retain these similarities for client agent queries without the manual mapping of different data sources achieving an average accuracy of 57% before leader negotiation and an average accuracy of 61% after leader negotiation. The best matching accuracy achieved in a single test is 79%. This is directly applicable to the Department of Defense (DOD) that possesses many systems, which share information that enables the military to achieve their objectives. The DOD created an environment called the Joint Battlespace Infosphere (JBI) to solve this integration problem. This research improves upon the JBI's use of exact matching of field names for integrating the information within the environment. It simulates this type of interaction by demonstrating agents wrapped around different databases negotiating and generating an ontology. An agent-generated ontology is compared with an expert generated ontology and testing uses a set of queries run against the ontologies show that this technique can be useful in a distributed information environment.