Issues in complex event processing: status and prospects in the Big Data era

Ioannis Flouris, Nikos Giatrakos, Antonios Deligiannakis, Minos Garofalakis, Michael Kamp, Michael Mock

Research output: Contribution to journalArticleResearchpeer-review

98 Citations (Scopus)

Abstract

Many Big Data technologies were built to enable the processing of human generated data, setting aside the enormous amount of data generated from Machine-to-Machine (M2M) interactions and Internet-of-Things (IoT) platforms. Such interactions create real-time data streams that are much more structured, often in the form of series of event occurrences. In this paper, we provide an overview on the main research issues confronted by existing Complex Event Processing (CEP) techniques, with an emphasis on query optimization aspects. Our study expands on both deterministic and probabilistic event models and spans from centralized to distributed network settings. In that, we cover a wide range of approaches in the CEP domain and review the current status of techniques that tackle efficient query processing. These techniques serve as a starting point for developing Big Data oriented CEP applications. Therefore, we further study the issues that arise upon trying to apply those techniques over Big Data enabling technologies, as is the case with cloud platforms. Furthermore, we expand on the synergies among Predictive Analytics and CEP with an emphasis on scalability and elasticity considerations in cloud platforms with potentially dispersed resource pools.

Original languageEnglish
Pages (from-to)217-236
Number of pages20
JournalJournal of Systems and Software
Volume127
DOIs
Publication statusPublished - May 2017
Externally publishedYes

Keywords

  • Cloud computing
  • Complex event processing
  • Predictive analytics

Cite this