Title: Agent Miner: An Algorithm for Discovering Agent Systems from Event Data

Abstract: The problem of process discovery in process mining studies ways to construct process models that encode business processes that induced event data recorded by IT systems. Most existing discovery algorithms are concerned with constructing models that represent the control flow of the processes. Agent system mining argues that business processes often emerge from interactions of autonomous agents and uses event data to construct models of the agents and their interactions. This paradigm shift from the control flow to agent system discovery proves beneficial when interacting agents have produced the underlying data. This paper presents an algorithm, called Agent Miner, for discovering models of agents and their interactions that compose the system that has generated the business processes recorded in the input event data. The conducted evaluation using our open-source implementation of Agent Miner over publicly available industrial datasets confirms that the approach can unveil insights into the process participants and their interaction patterns and often discovers models that describe the data more accurately in terms of precision and recall and are smaller in size than the corresponding models discovered using conventional discovery algorithms.