Citation Context

...ed in event logs by contemporary workflow management systems can be exploited for the purpose of risk identification. Various approaches for predicting timeliness have been proposed in the literature =-=[8, 9]-=- and serve as a starting point for our work. Van der Aalst et al.’s approach [8] builds an annotated transition system and remaining process time is then predicted based on the average of earlier case...

Citation Context

...oint. Let us consider Te as the remaining cycle time, i.e. the amount of time estimated to complete the current instance given Tc based on past executions, which can be computed using the approach in =-=[2]-=-. Then the probability of exceeding MCT can be computed as 1 − MCT /(Te + Tc) if Te + Tc > MCT and is equal to 0 if Te + Tc ≤ MCT . If this probability is greater than a tolerance value (e.g. 60%), we...

Citation Context

...rity. Recommendation concerns enacting the appropriate actions to prevent faults from occurring. We dealt with run-time risk detection in previous work [5]. Regarding prediction, van der Aalst et al. =-=[19]-=- propose an approach to predict the remaining cycle time till the end of the execution of a process instance on the basis of the process control-flow, while Folino et al. [10] use decision trees to im...

Citation Context

...the time differences between causally related events and compute basic statistics such as averages, variances, and confidence intervals. This way it is possible to identify the main bottlenecks [10], =-=[19]-=-. Information about resources can be used to discover roles, i.e., groups of people frequently executing related activities [20]. Here, standard clustering techniques can be used. It is also possible ...

Citation Context

...until the end of process instances on the basis of the current state, which consists of the number of executions of each process activity and the current values of process variables. It is similar to =-=[2]-=- when a multi-set abstraction is used, with, additionally, the current values of process variables are also taken into account. Dependent Characteristic: Remaining Time Until The End of the Case. Inde...

Citation Context

...maining cycle time of a case by applying non-parametric regression and using case data as predictor variables [20]. The approach for predicting remaining process time proposed by van der Aalst et al. =-=[18]-=- is based on building an annotated transitionsystem and estimating the average remaining time of cases that visited the same state previously. In contrast, our approach predicts the likelihood of cas...

Citation Context

....9). 6 Related Work In the literature, there are some works that provide approaches for generating predictions and recommendations during process execution and are focused on the time perspective. In =-=[3,2]-=-, the authors present a set of approaches based on annotated transition systems containing time information extracted from event logs. The annotated transition systems are used to check time conforman...

Citation Context

...oint. Let us consider Te as the remaining cycle time, i.e. the amount of time estimated to complete the current instance given Tc based on past executions, which can be computed using the approach in =-=[2]-=-. Then the probability of exceeding MCT can be computed as 1 −MCT/(Te + Tc) if Te+Tc > MCT and is equal to 0 if Te+Tc ≤ MCT . If this probability is greater than a tolerance value (e.g. 60%), we notif...