Selection Strategies in Complex Event Recognition

Author nameMaria Petropanagiotaki
TitleSelection Strategies in Complex Event Recognition
Year2017-2018
Supervisor

Alexandros Artikis

AlexandrosArtikis

Summary

This work is situated in the field of Complex Event Processing, where the aim is to detect specific patterns in a stream of data. We consider the CEP engine called Wayeb, which uses finite symbolic automata to that end and is written in Scala. Specifically, we expand its capabilities, by implementing two selection strategies: skill-till-next and skip-till-any. This allows us to detect events that are not contiguous in the stream, a capability that Wayeb lacked. Skip-till-next ignores irrelevant events to the given pattern, whereas skip-till-any finds all the possible matchings that contribute to a pattern formation. In order to implement these selections strategies, we use a unary operator so that we can detect events that are not of the operands form. Furthermore, we provide examples in string-recognition and marine traffic monitoring as practical implementation of our work.