Analisis dan Penerapan Process Mining Pada Data Perkuliahan Online Studi Kasus Virtual Class Universitas Lampung

Abstract

— Increasing data innovation collectively affects all regions of the association. The development of the Corona virus episode towards the beginning of 2020 has made all forms of movement carried out online. Lampung University is one of the institutions affected by the Covid-19 outbreak. The rampant transmission of this virus requires that all the academics of the Lampung campus complete internet learning until things return to normal. All learning exercises are diverted by utilizing an online-based framework called Virtual Class University of Lampung. This research uses process mining. Process mining is a strategy that applies unique calculations to record information. Process mining is a valuation strategy between cycle models and event or event log information contained in a data frame. Event Logs can be easily incorporated into process models using heuristic mining algorithms. The best cycle model is obtained by utilizing three constraints, namely Relative-to-best Threshold (RT), Positive Observations Threshold (PT), and Dependency Threshold (DT). This limit is used to find the best health value. Health values are used to demonstrate model cycles by logging true or false events. If they match, the bottleneck found is valid or actually occurred. In addition, for this situation study, the soundness value obtained indicates suitability so that the obstacles that occur can be found. Each subject has a different frequency for activities completed in the Virtual Class. Process mining is the act of creating a cyclical model from the log of events that occurred, planning to work in the most common way of identifying and solving a problem. For this case the event log used is the event log from Virtual Class University of Lampung and handle mining can help distinguish bottlenecks that occur in the selected course.