Cause vs. effect in context-sensitive prediction of business process instances

Predicting undesirable events during the execution of a business process instance provides participants with an opportunity to intervene and keep aligned its goals. Few approaches for tackling this challenge consider multi-perspective view, where flow perspective is combined surrounding context. Given many sources data in today's world, context can vary widely have various meanings. This paper addresses issue being cause or effect next event impact on prediction. We leverage previous work probabilistic models develop Dynamic Bayesian Network technique. Probabilistic are considered comprehensible they allow end-user his her understanding domain be involved Our technique attributes that either relationship towards event. evaluate our two real-life sets benchmark it other techniques from field predictive monitoring. The results show solution achieves superior prediction if information correctly introduced into model.