XQuery meets Datalog: Data Relevance Query for workflow trustworthiness

We propose data relevance query as a comprehensive way to help understand data relations built during execution of workflows. Correct data relation is critical for trusting workflow systems. The language RQL (Relevance Query Language) is designed for users to write relevance query declaratively. RQL combines Datalog into XQuery, such that it has the expressive control structures as XQuery and also has the deductive ability of Datalog. We use Java bytecode as a sample language to demonstrate how relevance facts can be captured. The captured relevance facts are then queried to understand how data are processed by bytecode programs.