Nothing Special   »   [go: up one dir, main page]

There is a newer version of the record available.

Published March 22, 2023 | Version 1.0
Dataset Open

Evaluation datasets and results of the paper "Can I Trust My Simulation Model? Measuring the Quality of Business Process Simulation Models"

  • 1. University of Tartu
  • 2. University of Toronto
  • 3. York University

Description

Datasets and files used in the evaluation of the publication entitled "Can I Trust My Simulation Model? Measuring the Quality of Business Process Simulation Models", where:

  • BPS-models/: folder containing the BPS models used in the evaluation (the BPS models discovered by ServiceMiner are not included due to privacy reasons).
    • The BPS models discovered by SIMOD are composed of i) a BPMN file with the process model structure, and ii) a JSON file with the parameters of the simulation. These files correspond to the format of Prosimos simulation engine (https://prosimos.cloud.ut.ee/).
    • The BPS models of the Loan Application process are composed of a BPMN file with both the process model structure and parameters, corresponding to the format of the BIMP simulator used in APROMORE (https://apromore.com/).
  • measures/: folder containing the distance values of each measure reported in the paper.
  • original-event-logs/: folder containing the (train and test) event logs used in the evaluation.
  • simulated-logs/: folder containing the simulated logs evaluated in the paper (synthetic, SIMOD, and ServiceMiner).
  • ComputeLogDistance.py: script to compute the distance measures proposed in the paper.

 

To evaluate the distance measures of a set of simulated event logs in the folder simulated_logs/ against the test log test_event_log.csv.gz, run:
    python ComputeLogDistance.py -cfld test_event_log.csv.gz simulated_logs/

*The flag -cfld is optional, due to the high computational complexity of the CFLD measure.

WARNING: set the column names of each log accordingly (where log_1_ids are the IDs of the test log, and log_2_ids the IDs of the simulated logs). Examples:

# Column IDs for the (train/test) real-life logs, and the SIMOD simulated logs.
EventLogIDs(
    case='case_id',
    activity='activity',
    start_time='start_time',
    end_time='end_time',
    resource='resource'
)

# Column IDs for the Loan Application simulated logs.
EventLogIDs(
    case='case_id',
    activity='activity',
    start_time='Start_Time',
    end_time='End_Time',
    resource='resource'
)

# Column IDs for the ServiceMiner simulated logs.
EventLogIDs(
    case='case_id',
    activity='Activity',
    start_time='start_time',
    end_time='end_time',
    resource='Resource'
)

 

Files

BPS-models.zip

Files (180.2 MB)

Name Size Download all
md5:08be63b77b0bf794b605d3f598440738
346.3 kB Preview Download
md5:c20b24adc511a7fbd66cac9daefee863
37.0 kB Download
md5:9d82a994a23342340af20038a2429a24
26.2 kB Preview Download
md5:9dcdff463723126c6369bcd5fe089f55
12.4 MB Preview Download
md5:446c99118b46ef82ddcdab81b83976f7
2.3 kB Preview Download
md5:35d215aa1d5278b4ecddef2e97e5a127
167.4 MB Preview Download

Additional details

Funding

PIX – The Process Improvement Explorer: Automated Discovery and Assessment of Business Process Improvement Opportunities 834141
European Commission