Scalable indexing algorithm for multi-dimensional time-gap analysis with distributed computing

Data has been become increasingly and abundantly available as industries have applied numerous information systems to automate business process execution. Process mining focuses on discovering knowledge from historical data repositories to support better decision making, mostly from a single perspective. Multi-perspective, or multi-dimensional process mining becomes an open issue in process mining working groups, since the nature of event logs vary according to its domain, and a single process model might not be justified. Notwithstanding the previous work and the multi-dimensional process mining approaches developed therein, the contents of iterative indexing method and platform-dependent computational issues cause problems on scalability and usability respecting real world implementation. In response to such problems, the present study formulated a scalable indexing algorithm for multi-dimensional process analysis with distributed computing. A new solution is applied wherein we index only attributes inside the selected events and show only a reduced graph of long-duration gaps between events. The implementation is done with an independent online analytical tool. Additionally, case study of an actual port is provided to illustrate and alidate our method.