Mining Recency–Frequency–Monetary enriched insights into resources’ collaboration behavior from event data

Organizations increasingly rely on teamwork to achieve their goals. Therefore they continuously strive to improve their teams as their performance is interwoven with that of the organization. To implement beneficial changes, accurate insights into the working of the team are necessary. However, team leaders tend to have an understanding of the team's collaboration that is subjective and seldom completely accurate. Recently there has been an increase in the adoption of digital support systems for collaborative work that capture objective data on how the work took place in reality. This creates the opportunity for data-driven extraction of insights into the collaboration behavior of a team. This data however, does not explicitly record the collaboration relationships, which many existing techniques expect as input. Therefore, these relationships first have to be discovered. Existing techniques that apply discovery are not generally applicable because their notion of collaboration is tailored to the application domain. Moreover, the information that these techniques extract from the data about the nature of the relationships is often limited to the network level. Therefore, this research proposes a generic algorithm that can discover collaboration relationships between resources from event data on any collaborative project. The algorithm adopts an established framework to provide insights into collaboration on a fine-grained level. To this end, three properties are calculated for both the resources and their collaboration relationships: a recency, frequency, and monetary value. The technique's ability to provide valuable insights into the team structure and characteristics is empirically validated on two use cases.