Student Registration Process Evaluation using Process Mining Case Study: IT Telkom Imelda Atastina 1 , Angelina Prima Kurniati 2 School of Computing, Telkom University Bandung, Indonesia imelda@telkomuniversity.ac.id 1 , angelina@telkomuniversity.ac.id 2 Abstract—The process of student registration in IT Telkom needs to be evaluated to enable the continual improvements. For this purpose, we use process mining to obtain a real overview of the actual process happens on students registration process through IT Telkom registration information system. Fuzzy mining approach is proven to be more suitable to be used in this case compared to Alpha algorithm, although it seems that the student registration process is a structured process. This paper contributes in presenting an experience of algorithm selection on creating an informative process model and recommendation on the student registration process improvement, especially for IT Telkom. Keywords— IT Telkom, student registration, fuzzy mining, alpha algorithm, process mining I. INTRODUCTION Telkom Institute of Technology (IT Telkom) is one of higher education institution that is located in Bandung, West Java, Indonesia. It has sizeable student body, which is about 6000 students. Telkom Institute of Technology (IT Telkom) has been using a registration information system for almost a decade. As a higher education institution who wants to implement information technology in supporting its business processes, IT Telkom needs to evaluate student registration process which is held in the beginning of each semester. This evaluation is aimed to maintain and improve the process performance. Process mining is a tool which can be used to evaluate the business process [8], [12]-[14]. This can be done due to the ability of process mining approach to obtain a model process describing the actual process, as recorded directly in the event log of the information system, which is registration information system in this case. Considering the availability of the event log of registration information system at IT Telkom, we decided to use process mining in this research. To obtain a good process model, the first step to do in this research is choosing the most suitable algorithm for the case [10]. It needs to be done because there are many algorithms which can be used in process mining technique, such as alpha algorithm [13], heuristics algorithm [15], fuzzy mining [5], etc. But there is no guidance on how to choose a suitable algorithm in a specific case to obtain a good process model. After getting a good process model, we can evaluate the registration process, and conclude the recommendation to improve performance of student registration process. In his book [13], Aalst states that a business process which is well-defined and having an event log is a structured process, or so-called “lasagna process”. This can be applied to student registration process of IT Telkom. But after implementing alpha algorithm using ProM, the resulted process model of student registration process is not informative. Only when we implement fuzzy mining approach, the process model obtained is logical and can be used as a reference for evaluation. This paper is structured in five sections: Section (1) provides an overview of the contents of the paper, Section (2) presents a summary of the theory being used in the study, Section (3) provides an overview of the process mining implementation using alpha and fuzzy mining algorithm and the results of its implementation, Section (4) contains the result analysis, discussion and recommendations based on an evaluation of the process models obtained, and Section (5) for the conclusions. II. THEORY A. Process Mining Process mining is a study of implementation of data mining techniques on business process [2], [7], [13]. Process mining techniques allow for extracting information from event logs. Extraction of information can be presented in model process. There are many kinds of process model, such as Petri Net, BPMN, YAWL, etc. [1], [3], [4], [8] For example, the audit trails of a workflow management system or the transaction logs of an enterprise resource planning system can be used to discover models describing processes, organizations, and products [4], [6], [7]. In general, the relationship between the model process, event logs, real business process can be seen in Figure 1.