Runtime Monitoring and Analysis of Influential Factors of Business Process Performance
Branimir Wetzstein, email@example.com, USTUTT
The approach deals with process performance management in the context of business processes that are implemented as WS-BPEL service compositions. The process performance is monitored at runtime in terms of a set of key performance indicators (KPIs). If monitoring shows that KPI targets are not met, machine learning techniques are used in order to learn the corresponding influential factors the KPIs depend on. Therefore, decision trees are constructed in an automated fashion based on history monitoring data and are presented to the user. Based on the learned influential factors, adaptation actions can be triggered in order to improve the process performance.
The Java-based prototype is based on the Apache ODE BPEL engine (http://ode.apache.org/). Monitoring is implemented using the ESPER event processing framework (http://esper.codehaus.org/) . The analysis part uses decision tree algorithms provided by the WEKA framework (http://www.cs.waikato.ac.nz/ml/weka/). A purchase order processing scenario (a BPEL process interacting with six Web services) has been implemented for evalauting the concepts.
Please send a request to contact person
Service Composition Layer, Process Monitoring, Business Intelligence, Data Mining
Prototype (for different scenarios)
Relationship with Future Internet and Internet of Services
<!!!! PLEASE COMPLETE !!!!>
Relationship with Cloud