A Framework for Human-Centered Exploration of Complex Event Log Graphs.

Graphs can conveniently model complex multi-relational characteristics. For making sense of such data, effective interpretable methods for their exploration are crucial, in order to provide insights that cover the relevant analytical questions and are understandable to humans. This paper presents a framework for human-centered exploration of attributed graphs on complex, i.e., large and heterogeneous event logs. The proposed approach is based on specific graph modeling, graph summarization and local pattern mining methods. We demonstrate promising results in the context of a real-world industrial dataset.