Text-Aware Predictive Monitoring of Business Processes

The real-time prediction of business processes using historical event data is an important capability modern process monitoring systems. Existing methods are able to also exploit the perspective recorded events, in addition control-flow perspective. However, while well-structured numerical or categorical attributes considered many techniques, almost no technique utilize text documents written natural language, which can hold information critical task. In this paper, we illustrate design, implementation, and evaluation a novel text-aware model based on Long Short-Term Memory (LSTM) neural networks language models. proposed take categorical, textual into account predict activity timestamp next event, outcome, cycle time running instance. Experiments show that outperform state-of-the-art simulated real-world logs containing data.