Business process mining from e-commerce event web logs: Conformance checking and bottleneck identification

Abstract

A range of advanced methods have been formulated and utilized in the efforts of improving the business processes in many enterprises. One impacting enhancement technique is to employ process mining algorithms as modeling and analysis tools in order to provide the actual business performance by digging the event log data and finding the useful information. This paper focuses on the applications of process mining in e-commerce industry. Event log data with timestamps were retrieved and analyzed from the web databases of an e-commerce company and process mining algorithms, like inductive miner and fuzzy miner were executed for generating the actual e-commerce business processes automatically and checking their conformance with the standardized processes as well as to early detecting any bottlenecks and issues in the e-commerce processes. Several e-commerce process issues were considered, such as item procurement, product order and delivery item tracking. The process mining modeling and its statistical results indicate that process mining can provide an efficient and effective tool for modeling and analyzing the e-commerce business processes allowing for real-time process auditing and reengineering.