Towards IoT-Driven Predictive Business Process Analytics

IEEE COINS 2020
International Conference on Omni-
layer Intelligent Systems
August 2020 - Online Presentation
Towards IoT-Driven Predictive Business Process
Analytics
Faculty of Computer Science and Engineering, Shahid Beheshti University, Tehran, Iran.
Department of Computer Engineering, Sharif University of Technology, Tehran, Iran.
Cyberspace Research Institute, Shahid Beheshti University, Tehran, Iran.
Speaker.
1
2
3
*
Erfan Elhami , Abolfazl Ansari , Bahar Farahani , Fereidoon Shams Aliee2 3* 11

2 / 22
TABLE OF CONTENTS
 INTRODUCTION
 Topic Area
 Data Driven Process Analysis
 IoT Driven PPM
 Process Event Log
 Case In Point
 Data Integration Methods
 RELATED WORK
 Previous Works Review
 Overview and Shortcomings
 PROPOSED APPROACH
 Proposed Approach Steps
 Building Blocks
 EXPERIMENTAL RESULTS
 Case Study
 Events Log and Implementation
 Results Evaluation
 CONCLUSION
 FUTURE WORK
 REFERENCES

3 / 22
INTRODUCTION - Topic Area
 Predictive Business Process Monitoring
• Subset of Process Mining
• Proactive Approach
• Uses Machine Learning (ML)
Business
Processes
AssetsPeople
Procedure
 Continuous Process Monitoring
 Business Process

4 / 22
INTRODUCTION – Data Driven Process Analysis
Business Process
Process Mining
Predictive Process
Monitoring (PPM)
Event Data
Process Data
Storage
Extract
Feed
Predictive Insight
Event Stream

5 / 22
INTRODUCTION – IoT Driven PPM
 Business processes are not isolated
from the environment
 Internet of Things (IoT)
IoT

6 / 22
INTRODUCTION – Process Event Log
 Process Data:
o Intrinsic Events: generated by performing process steps and
recorded as process event logs.
o Contextual Events: collected from the third-party process data
resources such as IoT devices and linked to the process,
indirectly.

7 / 22
INTRODUCTION – Case In Point
 Context-Aware PPM can have many applications in
the healthcare domain.
IoT
Healthcare Devices
Context-Aware
PPM of the
Healthcare Process
Process Data
Contextual Data (IoT)
Healthcare
Process

8 / 22
INTRODUCTION – Data Integration Methods
 Only a few works consider contextual
events in PPM approaches.
 No specific integration solutions

9 / 22
RELATED WORK – Previous Works Review
 The existing solutions and presented in the following five categories:
I. Time-Based Predictions,
II. Process Output Prediction,
III. Process Path Predictions,
IV. Process Risk Predictions,
V. Other Predictions and Works.
a. Prediction Output
b. Algorithms
c. Implementation Environment
d. Industry or Business Domain
e. Input Data Type
f. Context-Aware Approach
g. Incorporating IoT Data
h. Availability of The Dataset

10 / 22
RELATED WORK – Overview and Shortcomings
• Most Prediction: Time-Based Category (Remaining Time)
• Popular Algorithm: The Decision Tree
• New Trend: Using The ANN
• Implementation: ProM Framework
• ML Environment: Weka Toolkit
• Datasets: Limited Access
 Previous works: confirmed the importance of the contextual events in processes
 There is no:
o Specified proposed approach or architecture
o Specified steps of using contextual event data into a PPM framework
No serious attempt has been made to support transferring IoT events to the process analysis.
Only a few methods have been presented to integrate the context data with the process.

11 / 22
PROPOSED APPROACH - Proposed Approach Steps
 The basic idea of our approach is using IoT events as a process context
 The steps of the Proposed Approach are as follows:
1. Data Collection
2. Data Integration
3. Data Preprocessing
4. Data Processing
5. Presentation

12 / 22
PROPOSED APPROACH - Building Blocks
 The Building Blocks of the proposed context-aware PPM approach
 Inspired by Lambda Architectures

13 / 22
EXPERIMENTAL RESULTS - Case Study
 Case Study:
o The aircraft take-off process
o The simplified take-off process in terms of BPMN diagram
 Process Flow:
o Seven activities
o A decision point

14 / 22
EXPERIMENTAL RESULTS - Case Study
 IoT devices at the airport continuously collect the
changes in weather.
 The weather conditions as process context

15 / 22
EXPERIMENTAL RESULTS - Events Log and Implemention
 Synthetic event logs
 The aircraft take-off scenario and process rules
• Implemented in java.
 IoT events have been generated for three weather sensors
• Temperature,
• Wind speed,
• Humidity.

16 / 22
EXPERIMENTAL RESULTS - Events Log and Implementation
 Event logs containing 1000 cases and 7887 events.
 The event logs contain the process intrinsic information:
• Event ID,
• Case ID,
• Activity,
• Timestamp,
• Flight Number,
• Pilot Grade,
• And Plane Type
 The contextual events:
• Temperature,
• Wind Speed,
• Humidity.
We used 80% of the traces as the training set
and the remaining as the test set.

17 / 22
EXPERIMENTAL RESULTS - Events Log and Implementation
 Apache Kafka: Stream Processing Layer,
 Prediction model relies on the implementations of the ANN.
• Python’s Scikit-Learn
 Supervised Learning
 Feature vectors:
Pilot Grade, Aircraft Type, and Weather Information
 response variable:
Flight Permission and Correct Flight Plan labeled

18 / 22
EXPERIMENTAL RESULTS – Results Evaluation
 The evaluation metrics for the classifier
 Confusion Matrix
 Accuracy:
 Training set was 92%
 Test set was 91%
Flight Permission labeled 1 and Correct Flight Plan labeled 0

19 / 22
CONCLUSION
 Context-Aware PPM has many applications in different businesses and industries
 Limited amount of approaches for incorporate the contextual events into the PPM
 We mainly show primary steps for using IoT events in process prediction
 We propose to integrate the contextual events with the runtime process, for a
powerful correlation
 Building blocks for performing the proposed approach

20 / 22
FUTURE WORK
 More extensive architecture for using modern data
resources in PPM
 Real-world data set and more case studies

21 / 22
REFERENCES
[1] F. M. Maggi, C. Di Francescomarino, M. Dumas, and C. Ghidini, “Predictive monitoring of business processes,” in International conference on advanced information systems engineering, pp. 457–472,
Springer, 2014.
[2] R. Mans, M. Schonenberg, M. Song, W. M. van der Aalst, and P. J. Bakker, “Process mining in healthcare: a case study,” in conference; Healthinf 2008; 2008-01-28; 2008-01-31, pp. 118–125, INSTICC
Press, 2008.
[3] I. Verenich, Explainable predictive monitoring of temporal measures of business processes. PhD thesis, Tartu University, 2019.
[4] F. Koetter and M. Kochanowski, “A model-driven approach for eventbased business process monitoring,” Information Systems and e-Business Management, vol. 13, no. 1, pp. 5–36, 2015.
[5] M. Borkowski, W. Fdhila, M. Nardelli, S. Rinderle-Ma, and S. Schulte, “Event-based failure prediction in distributed business processes,” Information Systems, vol. 81, pp. 220–235, 2019.
[6] A. Bevacqua, M. Carnuccio, F. Folino, M. Guarascio, and L. Pontieri, “Adaptive trace abstraction approach for predicting business process performances.,” in SEBD, pp. 437–444, 2013.
[7] A. Bevacqua, M. Carnuccio, F. Folino, M. Guarascio, and L. Pontieri, “A data-driven prediction framework for analyzing and monitoring business process performances,” in International Conference on
Enterprise Information Systems, pp. 100–117, Springer, 2013.
[8] F. Folino, M. Guarascio, and L. Pontieri, “Context-aware predictions on business processes: an ensemble-based solution,” in International Workshop on New Frontiers in Mining Complex Patterns, pp.
215–229, Springer, 2012.
[9] B. F. Hompes, J. C. Buijs, and W. M. van der Aalst, “A generic framework for context-aware process performance analysis,” in OTM Confederated International Conferences” On the Move to Meaningful
Internet Systems”, pp. 300–317, Springer, 2016.
[10] E. Cesario, F. Folino, M. Guarascio, and L. Pontieri, “A cloud-based prediction framework for analyzing business process performances,” in International Conference on Availability, Reliability, and
Security, pp. 63–80, Springer, 2016.
[11] M. De Leoni, W. M. Van der Aalst, and M. Dees, “A general framework for correlating business process characteristics,” in International Conference on Business Process Management, pp. 250–266,
Springer, 2014.
[12] A. Cuzzocrea, F. Folino, M. Guarascio, and L. Pontieri, “Predictive monitoring of temporally-aggregated performance indicators of business processes against low-level streaming events,” Information
Systems, vol. 81, pp. 236–266, 2019.
[13] C. Di Francescomarino, M. Dumas, F. M. Maggi, and I. Teinemaa, “Clustering-based predictive process monitoring,” IEEE transactions on services computing, vol. 12, no. 6, pp. 896–909, 2017.
[14] M. De Leoni, W. M. van der Aalst, and M. Dees, “A general process mining framework for correlating, predicting and clustering dynamic behavior based on event logs,” Information Systems, vol. 56, pp.
235–257, 2016.
[15] M. Ceci, P. F. Lanotte, F. Fumarola, D. P. Cavallo, and D. Malerba, “Completion time and next activity prediction of processes using sequential pattern mining,” in International Conference on Discovery
Science, pp. 49–61, Springer, 2014.
[16] S. Huber, M. Fietta, and S. Hof, “Next step recommendation and prediction based on process mining in adaptive case management,” in Proceedings of the 7th International Conference on Subject-
Oriented Business Process Management, pp. 1–9, 2015.
[17] A. Rogge-Solti, L. Vana, and J. Mendling, “Time series petri net modelsenrichment and prediction,” CEUR Workshop Proceedings, 2015.
[18] M. Polato, A. Sperduti, A. Burattin, and M. de Leoni, “Data-aware remaining time prediction of business process instances,” in 2014 International Joint Conference on Neural Networks (IJCNN), pp.
816– 823, IEEE, 2014.
[19] S. Pandey, S. Nepal, and S. Chen, “A test-bed for the evaluation of business process prediction techniques,” in 7th International Conference on Collaborative Computing: Networking, Applications and
Worksharing (CollaborateCom), pp. 382–391, IEEE, 2011.
[20] A. Rogge-Solti and M. Weske, “Prediction of remaining service execution time using stochastic petri nets with arbitrary firing delays,” in International Conference on Service-Oriented Computing, pp.
389–403, Springer, 2013.
[21] Y. Liu, H. Zhang, C. Li, and R. J. Jiao, “Workflow simulation for operational decision support using event graph through process mining,” Decision Support Systems, vol. 52, no. 3, pp. 685–697, 2012.
[22] A. Senderovich, M. Weidlich, A. Gal, and A. Mandelbaum, “Queue mining for delay prediction in multi-class service processes,” Information Systems, vol. 53, pp. 278–295, 2015.
[23] G. T. Lakshmanan, S. Duan, P. T. Keyser, F. Curbera, and R. Khalaf, “Predictive analytics for semi-structured case oriented business processes,” in International Conference on Business Process
Management, pp. 640–651, Springer, 2010.

22 / 22
REFERENCES
[24] S. Pravilovic, A. Appice, and D. Malerba, “Process mining to forecast the future of running cases,” in International Workshop on New Frontiers in Mining Complex Patterns, pp. 67–81, Springer, 2013.
[25] I. Teinemaa, M. Dumas, F. M. Maggi, and C. Di Francescomarino, “Predictive business process monitoring with structured and unstructured data,” in International Conference on Business Process
Management, pp. 401–417, Springer, 2016.
[26] F. Folino, M. Guarascio, and L. Pontieri, “A prediction framework for proactively monitoring aggregate process-performance indicators,” in 2015 IEEE 19th International Enterprise Distributed Object
Computing Conference, pp. 128–133, IEEE, 2015.
[27] H. Horita, H. Hirayama, T. Hayase, Y. Tahara, and A. Ohsuga, “Process mining approach based on partial structures of event logs and decision tree learning,” in 2016 5th IIAI International Congress on
Advanced Applied Informatics (IIAI-AAI), pp. 113–118, IEEE, 2016.
[28] M. T. Wynn, W. Z. Low, A. H. ter Hofstede, and W. Nauta, “A framework for cost-aware process management: cost reporting and cost prediction,” Journal of Universal Computer Science, vol. 20, no. 3,
pp. 406–430, 2014.
[29] D. Breuker, M. Matzner, P. Delfmann, and J. Becker, “Comprehensible predictive models for business processes.,” Mis Quarterly, vol. 40, no. 4, pp. 1009–1034, 2016.
[30] C. Cabanillas, C. Di Ciccio, J. Mendling, and A. Baumgrass, “Predictive task monitoring for business processes,” in International Conference on Business Process Management, pp. 424–432, Springer,
2014.
[31] M. Unuvar, G. T. Lakshmanan, and Y. N. Doganata, “Leveraging path information to generate predictions for parallel business processes,” Knowledge and Information Systems, vol. 47, no. 2, pp. 433–
461, 2016.
[32] D. Breuker, P. Delfmann, M. Matzner, and J. Becker, “Designing and evaluating an interpretable predictive modeling technique for business processes,” in International Conference on Business
Process Management, pp. 541–553, Springer, 2014.
[33] M. Le, D. Nauck, B. Gabrys, and T. Martin, “Sequential clustering for event sequences and its impact on next process step prediction,” in International Conference on Information Processing and
Management of Uncertainty in Knowledge-Based Systems, pp. 168–178, Springer, 2014.
[34] J. Becker, D. Breuker, P. Delfmann, and M. Matzner, “Designing and implementing a framework for event-based predictive modelling of business processes,” Enterprise modelling and information
systems architectures-EMISA 2014, 2014.
[35] M. Le, B. Gabrys, and D. Nauck, “A hybrid model for business process event and outcome prediction,” Expert Systems, vol. 34, no. 5, p. e12079, 2017.
[36] R. Conforti, M. de Leoni, M. La Rosa, W. M. van der Aalst, and A. H. ter Hofstede, “A recommendation system for predicting risks across multiple business process instances,” Decision Support
Systems, vol. 69, pp. 1–19, 2015.
[37] A. Pika, W. M. van der Aalst, M. T. Wynn, C. J. Fidge, and A. H. ter Hofstede, “Evaluating and predicting overall process risk using event logs,” Information Sciences, vol. 352, pp. 98–120, 2016.
[38] R. Conforti, S. Fink, J. Manderscheid, and M. Roglinger, “Prism–a pre- ¨ dictive risk monitoring approach for business processes,” in International Conference on Business Process Management, pp.
383–400, Springer, 2016.
[39] N. Tax, I. Verenich, M. La Rosa, and M. Dumas, “Predictive business process monitoring with lstm neural networks,” in International Conference on Advanced Information Systems Engineering, pp.
477–492, Springer, 2017.
[40] W. M. Van der Aalst, M. H. Schonenberg, and M. Song, “Time prediction based on process mining,” Information systems, vol. 36, no. 2, pp. 450–475, 2011.
[41] A. Dohmen and J. Moormann, “Identifying drivers of inefficiency in business processes: a dea and data mining perspective,” in Enterprise, Business-Process and Information Systems Modeling, pp.
120–132, Springer, 2010.
[42] L. Zeng, C. Lingenfelder, H. Lei, and H. Chang, “Event-driven quality of service prediction,” in International Conference on Service-Oriented Computing, pp. 147–161, Springer, 2008.
[43] M. Hausenblas and N. Bijnens, “Lambda architecture [online: http://lambda-architecture.net/],” 2017, Accessed [2020-01].
[44] J. Becker, D. Breuker, P. Delfmann, and M. Matzner, “Designing and implementing a framework for event-based predictive modelling of business processes,” Enterprise modelling and information
systems architectures-EMISA 2014, 2014.
[45] M. De Leoni, W. M. van der Aalst, and M. Dees, “A general process mining framework for correlating, predicting and clustering dynamic behavior based on event logs,” Information Systems, vol. 56, pp.
235– 257, 2016.
[46] T. da Cunha Mattos, F. M. Santoro, K. Revoredo, and V. T. Nunes, “Formalizing the situation of a business process activity,” in Proceedings of the 2012 IEEE 16th International Conference on
Computer Supported Cooperative Work in Design (CSCWD), pp. 128–134, IEEE, 2012.

You can find me at:
erfan.elhami@gmail.com
linkedin.com/in/erfanelhami

Towards IoT-Driven Predictive Business Process Analytics

More Related Content

What's hot (16)

Similar to Towards IoT-Driven Predictive Business Process Analytics (20)

Recently uploaded (20)

Towards IoT-Driven Predictive Business Process Analytics