Automated rule generation for complex event processing

Abstract

The key concept of Complex Event Processing (CEP) is setting up accurate queries to pick up the important events. Since all CEP engine s support SQL like queries, this rule setup requires some technical skill s plus domain knowledge. The best solution to addres s this issue is to automate the query generation of CEP. Existing automated query generation methodologies are computationally expensive and are not fully automated process es . This study addresses the above two issues by proposing a shapelet based approach . This new approach is not computationally expensive, and it is a fully automated process with zero manual user intervention required. The proposed method uses the computationally efficient algorithm call ed Fast Shapelet Selection (FSS) algorithm. This FSS algorithm is used to extract the shapelets from data set. Then extracted shaplets are used to generate CEP queries. This proposed method can be used analyze to multivariant time series and this is more efficient than previously proposed shapelet based approaches.