Formalization and Detection of Events over a Sliding Window in Active Databases Using Interval-Based Semantics

Trend analysis and forecasting applications (e.g., securities trading, stock market, and after-the-fact diagnosis) need event detection along a moving time window. Event-driven approaches using a push-paradigm play a significant role in many real-world applications since changes detected are crucial for these applications. In active databases that provide push-paradigm, an event was defined to be an instantaneous, atomic occurrence of interest and the time of occurrence of the last event in an event expression was used as the time of occurrence for an entire event expression (detection-based semantics), rather than the interval over which an event expression occurs (interval-based semantics). Currently, all active databases detect events using the detection-based semantics rather than the interval-based semantics. This introduces semantic discrepancy for some operators when they are composed more than once. In this paper, we present the need for interval-based semantics for detecting events over a sliding window (or in continuous context) and formalize the semantics of Snoop (an event specification language) event operators using interval-based semantics.