Customer Journey and Win Back using Object-Centric Process Mining to Better Understand it from Multiple Viewpoints and Further Promoting Better Offers to Retain Customers

Contact

Name

Alessandro Berti

Software Engineer

Phone

work
+49 241 80 21949

Email

E-Mail
 

Customer Journey and Win Back using Object-Centric Process Mining to Better Understand it from Multiple Viewpoints and Further Promoting Better Offers to Retain Customers

Student: Ahmed ElDesouki

Title: Customer Journey and Win Back using Object-Centric Process Mining to Better Understand it from Multiple Viewpoints and Further Promoting Better Offers to Retain Customers

Supervisor: Alessandro Berti

1st Examiner: Prof. Wil M.P. van der Aalst

2nd Examiner: Prof. Dr. Ulrik Schroeder

Summary

Many companies seek to explore, debug and enhance their processes across their different departments. Those companies seek customer loyalty and satisfaction while spending less money. While there exists a plethora of research on process mining approaches, that tackle the classical event log from discovery to enhancing these processes. Real-life operational systems may suffer from convergence and divergence problems. The classical event log can not address those problems because of the event per case notion restriction. On the other hand, an object-centric event log can. In this work, we address and explore a real-life system. We are using object-centric process mining techniques, Specially organizational mining, on top of the object-centric event log. Celonis software initially suggested this approach, whereas we deliver more granular aggregations and support the union of network analysis multigraphs. Our work allows for more complex analysis and having multiple social metrics viewpoints over logs that help understands the organization better. We further evaluated this generic approach using Vodafone and other public process mining datasets. We helped with a more profound understanding of relationships between the resources from multiple viewpoints without jeopardizing the calculations of network analysis metrics, which would have happened in the case of the classical approaches.