On the Use of Knowledge Graph Completion Methods for Activity Recommendation in Business Process Modeling

Business process modeling is essential for organisations. However, it is a time-consuming task that requires expert knowledge. In particular, this is the case when modeling domain-specific processes, which often involves the consistent use of technical terminology. Process modelers can be supported through the provision of recommendations on how the model under development can be expanded. Activity recommendation is one such support approach, in which suitable activities to be inserted at a user-defined position are recommended. Recently, it has been suggested to treat activity recommendation as a knowledge graph completion task and to apply methods from this discipline. In this paper, we investigate different approaches to apply embedding- and rule-based knowledge graph completion methods out of the box and evaluate them in an experimental study. Additionally, we compare them to two methods that have specifically been designed for activity recommendation.

Automated summary

This paper investigates different approaches to apply embedding- and rule-based knowledge graph completion methods and compares them to methods specifically designed for activity recommendation.