ElogQP: An Event log Quality Pointer Tobias Ziolkowski 1 , Lennart Brandt 2 , Agnes Koschmider 1 1 Process Analytics Group, Computer Science Department, Kiel University, Germany {tzi|ak}@informatik.uni-kiel.de 2 stu113969@mail.uni-kiel.de Abstract. This paper presents ElogQP, a tool to detect data quality violations in an event log. Data quality issues significantly impact the process discovery result. Thus, ElogQP represents an essential step towards improved process discovery. Keywords: event log, process mining, data cleaning, imperfection patterns. 1 Introduction Event log files are used as input to any process mining algorithm aiming to discover an as-is process model or to identify bottlenecks. Although recently process mining has gained an impressive uptake, still, data quality violations often hamper the direct applicability of process mining techniques on an event log. There are several reasons for data quality violations like those that the recorded event data is not saved in the correct order, data entries are missing (e.g. timestamps or case ID) or are not recorded correctly (e.g. incomplete activity names). These quality violations lead to inappropriate event logs and finally significantly impact the process discovery result. To counteract data quality issues in process mining several approaches exist [1, 2, 3] like to define maturity levels for data quality [1], to use a framework of timestamp imperfections [2] or a framework for event log quality [3]. Better understanding of how data quality issues affect the event log quality led to the definition of so-called event log imperfection patterns [4]. This paper presents the Event log Quality Pointer (EloqQP) tool aiming to detect data quality violations. The tool allows to detect event log imperfection patterns and to classify the data violations according to data quality levels as specified in the process mining manifesto [5]. Beside this, a comparison between two event logs with respect to data quality violations is supported. Thus, ElogQP detects missing start or end activities and activities with incorrect order. Fig. 1 shows how ElogQP works when two event logs are used as input. The event log on the left-hand side is (more) complete, while on the right-hand side one timestamp and one activity are missing. When parsing both event logs, ElogQP returns data types that have been identified as data quality violations with a descriptive comment to understand the violation (see “Output of ElogQP”). The paper is structured as follows. Section 2 gives an overview of ElogQP. It describes the components and the functionality of the tool. Section 3 concludes the paper. J. Manner, S. Haarmann, S. Kolb, N. Herzberg, O. Kopp (Eds.): 13 th ZEUS Workshop, ZEUS 2021, Bamberg, held virtually due to Covid-19 pandemic, Germany, 25-26 February 2021, published at http://ceur-ws.org Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).