Information Systems version

Implementation of the technique paper submitted to Information Systems under the title "Unveiling the Causes of Waiting Time in Business Processes from Event Logs" by Katsiaryna Lashkevich, Fredrik Milani, David Chapela-Campa, Ihar Suvorau, and Marlon Dumas.

The technique takes as input an event log, in csv format wherein each row represents the execution of an activity (with start and end timestamps), and produces a report with the waiting time of each activity transition (e.g. from activity A to activity B) classified into five different categories.

• Waiting time due to batch processing: the activity instance is waiting until a set of instances of the same activity are accumulated and executed together.
• Waiting time due to resource contention: the resource that has to perform this activity is busy processing another activity.
• Waiting time due to prioritization: the resource that has to perform this activity is busy because they prioritized another activity that was supposed to be executed later.
• Waiting time due to resource unavailability: the resource is not working at that specific moment (e.g. non working weekend).
• Waiting time due to extraneous factors: the activity instance waits because of causes unrelated with the available data (e.g. the resource is busy working in another project, or the activity has to wait for an external event to happen).

Publication Evaluation

Synthetic Evaluation

The event logs used to evaluate "EQ1 - To which extent is the technique able to detect the presence or absence of certain waiting time causes?" can be found in this folder, under the names LoanApp_******.csv. The file LoanApp-names-mapping.txt contains the mapping between each of the csv log files and its identifier used in the paper.

The event logs used to evaluate "EQ2 - To what extent is the technique able to correctly quantify the amount of waiting time waste per each cause?" can be found in this folder, under the names manual_log_*.csv. The exact expected waiting times corresponding to each of these logs are in the files under the names manual_log_*_expected.csv.

The script file tests/test_infsys.py contains the code necessary to run these experiments.

poetry run pytest ./tests/test_infsys.py

Real-life Evaluation

The event log used for the real-life evaluation, corresponding to a production process, can be found in this folder, under the name Production.csv.