Learning & Reasoning for Logic-Based Complex Event Recognition

  • Alexander Artikis
  • Nikos Katzouris

Institute of Informatics, National Center for Scientific Research (NCSR) “Demokritos”, Athens, Greece
University of Pireaus, Pireaus, Greece


Complex Event Recognition (CER) refers to the detection of special events of interest in a multitude of large, heterogeneous and interdependent data streams.
In this short paper we discuss several aspects of CER research, such as efficient detection of patterns on event streams, handling uncertainty and noise in streams, and machine learning techniques for inferring interesting patterns.
We focus on logic-based methods for CER, which are of special interest thanks to their formal semantics and their direct connections to uncertainty handling and machine learning via Statistical Relational Learning and Inductive Logic Programming.