Online communities have become increasingly popular sources of information for both users and organisations. Every day thousands of users ask questions on these platforms, yet this knowledge-sharing process is not very studied. In this paper we aim to fill this knowledge-gap, by providing a general framework for studying the knowledge-sharing processes in such online communities. Specifically, we provide a three-step algorithm, that can create process models from interleaved and unlabelled conversations. We provide an instantiation of our framework, and conduct several experiments to evaluate its performance using the process mining tool Disco. From these experiments we show that it is possible to gain meaningful insights from the conversations on online communities using process mining techniques.
Published in Proceedings of the International Joint Conference on Neural Networks (IEEE WCCI IJCNN 2020); Glasgow, Scotland, UK; July 19-24, 2020.