Remaining Time Prediction for Collaborative Business Processes with Privacy Preservation

Jian Cao, Chi Wang, Wei Guan, Shiyou Qian, Haiyan Zhao

Abstract: In collaborative business processes that involve multiple organizations, privacy concerns prevent organizations from sharing the raw data of their activities. This makes it challenging to predict remaining time without access to data on completed activities of other organizations. To address this challenge, this research proposes a strategy for predicting remaining time in collaborative business processes, which involve sequential sub-processes executed by different partners, while preserving the privacy of organizations. The proposed strategy involves transferring latent information from precedent sub-processes to the models of latter sub-processes, rather than raw data. Two models were designed to implement this strategy, and the experimental results indicate that the prediction accuracy of the models is comparable to that of models that use raw data.