AIMS: An agent-based information management system in JBI-like environments

Abstract

One of the challenges for many mission-critical applications is how to determine the quality of information from distributed and heterogeneous data sources. In this paper we consider joint battlespace infosphere (JBI) or JBI-like collaborative information sharing environments, where the quality of published information often depends on the quality of the contributing information and of the sources that publish them. However, as the environment becomes larger and more diverse, it is becoming increasingly difficult for human operators to assess the quality of information from various data sources. To address this challenge, we develop AIMS, an agent-based information management system, to manage the quality of information in JBI-like environments, e.g., information trustworthiness. In our approach, each operator is associated with a software agent, called client agent. The client agent enables its operator to interact with the JBI repository via the services of query, publish, and subscribe. Moreover, the client agent collaborates with other agents to assess and learn the trustworthiness of information and the reliabilities of corresponding data sources from its pedigree and the feedback from operators.