A Logic-based Language for Data Streams

Data Stream Management Systems (DSMS) have attracted much interest, and various extensions of relational-database query languages have been proposed for data streams. However, relational query languages were built on the solid bedrock of logic, while current DSMS languages and their computation models are missing such foundations. In this paper, we show that continuous queries can be characterized using the familiar concepts of closed-world and local stratification, leading to Streamlog that allows a freer and more natural usage of nonmonotonic constructs than Datalog. Thus, Streamlog takes the query languages of DSMS to new levels of expressive power and removes the limitations that severely impair current commercial systems and research prototypes.