Mining Valuable Collaborations from Event Data Using the Recency-Frequency-Monetary Principle

Abstract

Collaborative work leads to better organizational performance. However, a team leader's view on collaboration does not always match reality. Due to the increased adoption of (online) collaboration systems in the wake of the COVID pandemic, more digital traces on collaboration are available for a wide variety of use cases. These traces allow for the discovery of accurate and objective insights into a team's inner workings. Existing social network discovery algorithms however, are often not tailored to discover collaborations. These techniques often have a different view on collaboration by mostly focusing on handover of work, resource profile similarity, or establishing relationships between resources when they work on the same case or activities without any restrictions. Furthermore, only the frequency of appearance of patterns is typically used as a measure of interestingness, which limits the kind of insights one can discover. Therefore we propose an algorithm to discover collaborations from event data using a more realistic approach than basing collaboration on the sequence of resources that carry out activities for the same case. Furthermore, a new research path is explored by adopting the Recency-Frequency-Monetary (RFM) concept, which is used in the marketing research field to assess customer value, in this context to value both the resource and the collaboration on these three dimensions. Our approach and the benefits of adopting RFM to gain insights are empirically demonstrated on a use case of collaboratively developing a curriculum.