Leveraging XML Search and Business Rules For Highly Configurable Data-Driven Decisions Support Services

Abstract

In this paper we present a multiple source data fusion software architecture designed to support data-driven decision support applications in a service oriented environment. This architecture addresses the requirement for instances of a single application to consume and combine multiple data sources and apply decision logic with the data and logic unknown when the application was developed. This architecture was developed for a Space Situational Awareness application where multiple data sources are used to characterize situations for space assets. The goal of this development was to overcome the difficulties associated with changes for new data sources coming online continuously, changes in assets, and evolution in complex and idiosyncratic data fusion logic. The data parsing, pre-processing, decision logic, data association with existing tracks, probability based situation characterization, and reporting are all orchestrated by a “fusion skeleton” that follows the architectural framework of the Dual Node Network technical architecture for Data Fusion & Resource Management (DF&RM). This application has been applied to scenario data built for multiple source fusion development and integration testing and has performed well in comparison to alternative fusion techniques.