Approach for Defining Rules in the Context of Complex Event Processing

Abstract

The vision of Industrie 4.0 and the Internet of Things (IoT) is based on the connection of smart products and smart machines equipped with sensors and actuators. The digitalization of industrial processes leads to the production of data streams. In this context, real-time analytics is becoming more and more important for business applications as a result of the need to deal with the growth of data and to react instantly to changes in the data streams. Complex event processing (CEP) is an efficient methodology to enable processing and real-time analysis of streams of data. The main focus of CEP is the detection of patterns in data streams. Therefore, a set of rules has to be predefined. These rules are characterized by various parameters. Defining the optimal values for these parameters is challenging. In current CEP systems, experts have to define the rule patterns. In this paper we suggest three ways to define rules: manual by domain experts, semi-automated by rule mining, or optimization. However, not all of these three ways can be applied to a production scenario or use case. Thus, we compare these approaches and match them with the appropriate production scenario.