Temporal Datalog Materialisation for Composite Event Recognition

A new arXiv paper proposes efficient materialisation for Temporal Datalog with stratified negation and no future dependencies, applied to composite event recognition. The authors map fragments of event specification languages into Temporal Datalog to enable comparison and stream reasoning. They introduce Streaming Trigger Graphs, extending state-of-the-art Datalog materialisation techniques for high-velocity event streams. The work addresses timely detection of critical situations like safety threats.