Article Preview
TopERP systems support business processes throughout the enterprise, with each of those processes consisting of separate but related transactions. Researchers have explored applying data mining to system logs for identifying the composition of transactions into processes within desktop and web-based enterprise applications (e.g. Greco, Guzzo, and Sacc, 2005; Khasawneh and Chan, 2006; Bayir, Toroslu, Cosar, and Fidan, 2009; Shen, Fitzhenry, and Dietterich, 2009; van der Aalst, 2011). Rozinat and van der Aalst (2008) have shown that process sequences mined from ERP system logs often deviate from prescribed processes. Discovering the ways in which users actually perform processes with systems can provide critical input to organizations, which rely on ERP systems for standardizing around best practices. At the same time, knowledge of typically performed processes enables a variety of system enhancements, such as intelligent workflow assistants, auto-generated To-Do lists, and the automating of workflows.