Go to the top

Welcome to
Process-Aware Artificial Intelligence Lab

Professor Marco Comuzzi

About

스크린샷 2020-11-19 오전 4.50.09
  • .
  • Office: UNIST Main Campus - 112 - 301

Introduction of Lab (Since 2016 ~ . )

This lab was established in 2016. It focuses on the application of machine learning techniques (mainly classification) to the analysis of business process event logs. As a lab, we are always trying to contribute to the main international academic conferences in the field: Int. Conf. on Business Process Management, Int. Conf. on Process Mining and Int. Conf. on Advanced Information Systems Engineering. Our research has always practical relevance. For instance we have applied the results of our research on event logs obtained from the Ulsan Port Authority and from a large university hospital in Korea.

Research Keywords

Business process management, process mining, machine learning, classification, anomaly detection.

Research Topics

  • Predictive monitoring of business processes using event logs
  • Anomaly detection in business process execution using event log
  • Explainable business process predictive monitoring
  • Quality analysis and improvement of event logs
  • Industrial and practical applications of business process mining

Members

Professional Profile

  • Head of Lab

마르코
Marco Comuzzi
Associate professor in School of Management Engineering (UNIST)
Email
mcomuzzi@unist.ac.kr
Office
UNIST Main Campus 112 - 301-9
Tel
+82 - 52-217-3187 (Fax: 3101)
Personal page
https://sites.google.com/site/marcocomuzziphd
  • Bio info
  • Since February 2016 : Assistant professor of Business Process Management with the School of Management Egineering, UNIST, Ulsan, Korea.
  • Lecturer in Business Process Computing and Governance at City University London (2013 – 2016).
  • Assistant Professor at the School of Industrial Engineering, Eindhoven University of Techonology (2009 – 2013).
  • Research fellow at Department of Computing, City University London (2008).
  • PhD in Computer Science Engineering at Politecnico di Milano, Milan, Italy (2007).

 

Click here to read about Professor Marco Comuzzi

  • Members

KakaoTalk_20190318_152957959
이종엽
MS program
Lab
UNIST Main Campus 112 - 302-2
Email
belllight@unist.ac.kr
  • Bio info
  • 2019 – present : Master program student at Industrial Engineering (Management Engineering), UNIST
  • 2014 – 2019 : BS in Management Engineering at UNIST
KakaoTalk_Photo_2020-11-14-22-07-13
권나현
MS program
Lab
UNIST Main Campus 112 - 302-2
Email
eekfskgus@unist.ac.kr
  • Bio info
  • 2021 – present : Master program student at Industrial Engineering, UNIST
  • 2016 – 2020 : BS in Industrial Engineering at UNIST
KakaoTalk_20190318_160701706
이수환 (Suhwan Lee)
MS program
Lab
UNIST Main Campus 112 - 302-2
Email
ghksdl6025@unist.ac.kr
  • Bio info
  • 2019 – present : Master program student at Industrial Engineering (Management Engineering), UNIST
  • 2012 – 2018 : BA in Management at UNIST
  • 2018 : Research internship program at IEL lab
KakaoTalk_20190318_145438092
김종찬 (Jongchan Kim)
Combined MS/PhD program
Lab
UNIST Main Campus 112 - 302-2
Email
jckim@unist.ac.kr
  • Bio info
  • 2016 – present : Combined MS/PhD program student at Industrial Engineering (Management Engineering), UNIST
  • 2011 – 2015 : BS in Management Engineering at UNIST
홈피사진
고종현 (Jonghyeon Ko)
Combined MS/PhD program
Lab
UNIST Main Campus 112 - 302-2
Email
whd1gus2@unist.ac.kr
  • Bio info
  • 2016 – present : Combined MS/PhD program  student at Industrial Engineering (Management Engineering), UNIST
  • 2012 – 2015 : BS in Statistics at Soong-Sil University (Double Major with Industrial Engineering)

 

  • Former Members

 

KakaoTalk_20190204_150109454
Marcela Vargas Santamaria
MS program
Email
alexmirker@gmail.com
  • Bio info
  • 2017 – 2018 : MS in Management Engineering at UNIST
  • 2017 – 2018 : Master program student in IEL lab
  • 2012 – 2016: B.B.A in Marketing and Internatioanl Business, Technology Management/Information Systems/Entrepreneurship and Finance/Accountiong at UNIST
_DSC0118
Erdenekhuu Unurjargal
MS program
Email
smile.eegii@gmail.com
  • Bio info
  • 2017 – 2018 : MS in Management Engineering at UNIST
  • 2017 – 2018 : Master program student in IEL lab
  • 2010 – 2014: BS in Technology Management/Information Systems/Entrepreneurship and Finance/Accountiong at UNIST
KakaoTalk_20160927_125956680
Hoang Nguyen
MS program
Email
hoangnguyen3892@unist.ac.kr
  • Bio info
  • 2016 – 2017 : MS in Management Engineering at UNIST
  • 2016 – 2017 : Master program student in IEL lab
  • 2010 – 2014: BS in Corporate Finance at Danang University of Economics
Image
Gyunam Park
Undergraduate intern
Email
shot9448@unist.ac.kr
  • Bio info
  • 2011 – 2016 : BS in Management/Computer science at UNIST
  • 2016 : Research internship program at IEL lab

Research

Recent works

  • Event Log Reconstruction Using Autoencoders
  • Remaining time prediction in outpatient process of Pusan National University Hosipital
  • Predicting next activity and attibutes in event-log
  • Time-series application on event-log
  • Business Process Intelligence Challenge 2019 (BPIC 2019)

Publications

Predictive monitoring using event logs

B. Tama, M. Comuzzi and J. Ko (2020) “An empirical investigation of different classifiers, encoding and ensemble schemes for next event prediction using business process event logs”, ACM Transactions on Intelligent Systems and Technology (accepted) 

B. Tama and M. Comuzzi (2019) “An empirical comparison of classification techniques for next event prediction using business process event logs”, Expert Systems with Applications, 129, 233-245.

M. Comuzzi, J. Ko and S. Lee (2019) “Predicting outpatient process flows to minimise the cost of handling returning patients: A case study”, Workshop on Process-Oriented Data Science for Healthcare (PODS4H 2019) held in conjunction with BPM 2019 (accepted)

M. Comuzzi, A. E. Marquez-Chamorro, and M. Resinas (2018) “A Hybrid Reliability Metric for SLA Predictive Monitoring”, 34th ACM Symposium on Applied Computing (accepted)

M. Comuzzi, A. Marquez-Chamorro and M. Resinas (2018) “Does your accurate process predictive monitoring model give reliable predictions?” 1st ICSOC Workshop on AI and Data Mining for Services.

 

Event log anomaly detection

J. Ko and M. Comuzzi (2020) “Detecting anomalies in business process event logs using statistical leverage”, Information Sciences (accepted)

J. Ko, J. Lee, and M. Comuzzi (2020) “AIR-BAGEL: An Interactive Root cause-Based Anomaly Generator for Event Logs”, 2nd Int. Conf. on Process Mining (ICPM) – Demonstration Track (accepted)

H. Nguyen, S. Lee, J. Kim, J. Ko and M. Comuzzi (2019) “Autoencoders for Improving Quality of Process Event Logs”, Expert Systems with Applications, 131, 132-147.

H. Nguyen and M. Comuzzi (2018) “Event log reconstruction using autoencoders”, 1st ICSOC Workshop on AI and Data Mining for Services.

 

 

Others

M. Comuzzi, C. Cappiello and G. Meroni (2020) “On the Need for Data Quality Assessment in Blockchains”, IEEE Internet Computing (accepted)

D. Beverungen et al. (2020) “Seven Paradoxes of Business Process Management in a Hyper-Connected World”, Business & Information Systems Engineering (accepted)

M. Vargas and M. Comuzzi (2020) “A multi-dimensional model of Enterprise Resource Planning Critical Success Factors”, Enterprise Information Systems, 14(1), 38-57.

B. Tama, M. Comuzzi, and Rhee, K.-H. (2019) “TSE-IDS: A Two-Stage Classifier Ensemble for Intelligent Anomaly-based Intrusion Detection System”, IEEE Access, 7, 94497-94507.

M. Comuzzi, E. Unurjargal, and C.H. Lim (2018) “Towards a design space for blockchain-based system reengineering”, 1st Workshop on Blockchains for Inter-Organizational Collaboration (BIOC’18), in conjunction with CAiSE 2018, pp. 138-143

M. Cho, M. Song, M. Comuzzi, and S. Yoo (2017) “Evaluating the effect of best practices for business process redesign: an evidence-based approach based on process mining techniques”, Decision Support Systems, 104, 92-103.

 

 

 

For a complete of Prof. Comuzzi’s publications:  Link

Projects

항만공사1


Project title : Development of a platform for process mining in port logistics (2019.05 ~ )

-프로세스 마이닝 용 항만공사 플랫폼 개발

Extended : Development of port logistics optimization service using big data-based container terminal congestion and import/export volume prediction system (2020.09 ~)

-빅데이터 기반 컨테이너 터미널 혼잡도·수출입 물동량 예측 시스템을 활용한 항만물류 최적화 서비스 개발
 

  • Company : Ulsan Port Authority
  • The objective of this project is to develop a platform for process mining in port logistics. While, in fact, process mining has been extensively applied in domains such as health care and public administration, applications of process mining in logistics and, particularly, port logistics, remain limited. This implies that the logistic industry is lagging behind on the potential savings and operational improvements normally accrued by the application of process mining techniques in real world settings. Therefore, a thorough analysis of the potential of process mining applied to the analysis of port logistics data is needed. This platform aims at identifying critical aspects concerning event log collection and analysis in (port) logistics, critical use cases of process mining in different reference business processes in (port) logistics, and demonstrating the effectiveness of different process mining techniques using real world port logistics data publicly available and, possibly, collected from Ulsan Port Authority (UPA). As far as process mining techniques are concerned, the focus will be mainly on predictive process monitoring techniques, an emerging branch of process mining that aims at adapting data mining techniques for predicting aspects of interests in business processes using event log data.

 

 

 


부산대


Project title : Development of a Hospital System Optimization Platform: Operating Room Scheduling and Outpatient Service Process Optimization (2018.10 ~ 2019.02)

-병원 시스템 최적화 플랫폼 개발: 수술실 스케줄러 및 최적의 외래 서비스 프로세스 개발을 중심으로

Extended : Development of a Outpatient Waiting Time Prediction Model and Improvement of Outpatient Service Processes (2019.10 ~ 2020.07)

-외래환자 대기시간 예측 모델 개발 및 외래 서비스 프로세스 개선
 

  • Company : Pusan National University Hospital
  • The objective of this project is to derive optimal outpatient service process and identify additional hospital system optimization issues. The concrete list is as follows : Identify outsourcing service process through process mining; Diagnosis and simulation of outpatient service processes; Derivation and verification of reduction of waiting time for medical treatment through simulation analysis; Improving the quality of event logs based on machine learning using data currently held by Pusan ​​National University Hospital and additional collected data; Development of Real-Time Waiting Time Prediction Model; Try to estimate and minimize the number of days spent in a particular patient group; Develop predictive machine learning model to predict remaining time using patient logs.

 

 


1


Project title : Blockchain-based System Engineering (2018.01 ~ 2019.12)

-블록체인 기반 시스템 구축

 

  • UNIST Research Find (Excellent research idea discovery project)
  • Although blockcain is recognised as a potentially disruptive technology for all industries, deployment of blockchain-based solutions outside finance and, specifically, cryptocurrencies and related applications, is still largely experimental. This is particularly true when blockchain is seen as a tool for engineering information exchange among a set of collaborating parties, or a “business network”. Many emerging applications are showing that there are a variety of design choices available while (re-)engineering a system using blockchain, such as the size and content of blocks, consensus rules to create new blocks, or the nature and shape of the smart contracts to be embedded in blocks. However, a scientific approach to exhaustively identify these design choices and guide designers during implementations is currently lacking.
    This project aims at filling this gap, by devising a methodology for reengineering systems using the blockchain technology. This methodology will comprise an analysis of the blockchain-based system design space, ie., the set of design choices available to designers, and a set of guidelines to support designers of a system during implementation.

 

 

 


인터블록체인


Project title : Blockchain Platform with Business Models towards Cross-domain Interoperability (2018.06 ~ 2020.02)

-크로스 도메인 호환성을 위한 블록체인 플랫폼 및 비즈니스 개발

 

  • IBRC POSTECH (대학ICT연구센터육성지원사업)
  • The Intelligent Enterprise Lab at UNIST, chaired by Prof Marco Comuzzi, conducts research in the areas of enterprise systems, business process management and blockchain-based systems design. The Lab has received a seed funding from UNIST for the year 2018 to develop a methodology for blockchain-based system design using case studies of real blockchain-based application development in the UNIST campus. The expertise developed within this seed funding research will constitute fundamental background knowledge for the research to be conducted in this project. In particular, the developed methodology will be extended to the case of interoperability of multiple blockchains. The Lab research in business process management also contributes to build important background knowledge for this project. Prof Comuzzi has a strong track record at UNIST and his previous research in the area of cross-organisational business process management modelling and enactment. This project represents an opportunity to transfer this knowledge into the context of blcockhain-based system interoperability design and enactment.

 

 

 


태성

Project title : Development of Data Analytics Methods to Identify the Sources of Oder (2017.10 ~ 2018.03)

-악취 원인 규명을 위한 데이터 분석 기법 연구

 

  • Company : TAE SUNG ENVIRONMENT INSTITUTE
  • The project objectives of this project are to be compressed into the following three. (1) Establish a big data integration platform related to odor. (2) Development of artificial intelligence for odor analysis. (3) Design of Ulsan Metropolitan City Odor Monitoring Service System based on Artificial Intelligence. Among them, the pure research goal excluding the platform and the design part of the system is related to the second business objective ‘deodorization artificial intelligence development’. Specifically, it is necessary to develop a high-performance odor recognition artificial intelligence model by identifying and solving various problems arising in the data collection-processing-analysis stage through the repeated experiment process of applying and testing various algorithms on odor data. It is a goal.

 

 

 


이벤트로그

Project title : Development of techniques for improving quality of event logs for process mining (2017.09 ~ 2019.08)

-프로세스 마이닝 용 이벤트로그 품질 제고 기술 개발

 

  • National Research Fund
  • The objective of this project is to design and implement new techniques to improve the quality of process event logs and to study how they impact the quality of process mining outcomes.

Year 1: To revise the literature and develop preliminary models of event log cleaning and imputation techniques.

Year 2: To develop event log cleaning and imputation techniques and assess their impact on the quality of process mining outcomes.

Year 3: To evaluate the developed techniques in real world cases and refine them using the feedback collected from practitioners.

 

 

 


Teaching

Just My Classes

Business Process Management

& Process Mining
After attending this course you should have achieved the following objectives:

  • Know the importance and relevance of Business Process Management (BPM) in modern organisations;
  • Know the phases of the BPM lifecycle;
  • Can identify the relevant processes to be considered in a BPM initiative
  • Can model a business process using BPMN (Business Process Modelling Notation)
  • Can analyse and redesign a business process to improve efficiency, possibly using simulation techniques
  • Know how business processes can be automated
  • Know the importance and role of ERP system to support organizational operations and business processes
  • Know the importance of Process Mining as a technique to discover and analyse business process
  • Can discover and analyse processes mined from existing enterprise systems event logs

Business Process Management

& Process Mining
After attending this course you should have achieved the following objectives:

  • Know the importance and relevance of Business Process Management (BPM) in modern organisations;
  • Know the phases of the BPM lifecycle;
  • Can identify the relevant processes to be considered in a BPM initiative
  • Can model a business process using BPMN (Business Process Modelling Notation)
  • Can analyse and redesign a business process to improve efficiency, possibly using simulation techniques
  • Know how business processes can be automated
  • Know the importance and role of ERP system to support organizational operations and business processes
  • Know the importance of Process Mining as a technique to discover and analyse business process
  • Can discover and analyse processes mined from existing enterprise systems event logs

Applied Programming for Management Engineering

The main aim of this course is to understand how a computer can store and process data effectively. To do this, data structures and algorithms must be studied with the help of a programming language. In this course, we will use Python as a programming language.

So, besides becoming an expert about data structures and algorithms, a (positive) side effect of this course is that you will learn a new programming language, i.e., Python (which, by the way, is used extensively in Data Science and scientific programming)

After attending this course you should have achieved the following objectives:

  • Understand growth rate of algorithms using the Big-Oh notation
  • Understand how recursion can be used to design algorithms
  • Understand the principles and usage application of basic data structures, such as arrays, stack, queues and lists.
  • Understand the principles and usage applications of advanced data structures, such as trees and hash tables
  • Understand and apply typical sorting and searching algorithms
  • Understand graphs, graph traversal and paths within graphs
  • Implement basic and advanced data structures in Python
  • Implement the algorithms covered by this course in Python
  • Design and evaluate efficient algorithms in Python to solve specific applied problems
  • Understand and utilise Python libraries to solve applied problems, such as using APIs (Twitter and Google maps are used as examples), creating and processing graphs, data science with Pandas.

Data-Driven Process Management

The objective of this course is to discuss the fundamentals of modeling, analyzing and improving business processes in organizations using the data produced by their execution. The first part of the course covers the basics of business process management (modeling, simulation, improvement); the second part of the course covers the basics of “process mining”, i.e., a collection of techniques to analyze process data.

Blockchain-based system engineering

The objective of this course is to cover the basics of blockchain technology as a technology for designing and implementing cross-organisational information systems. The course starts with an overview of blockchain technology and its emergence in the field of cryptocurrency and then will focus more extensively on designing systems using blockchain.
Particular attention will be given to Bitcoin (as the first and prominent example of public blockchain), Ethereum (for smart contracts), and Hyperledger (as an example of permissioned business blockchain).

The course requires some basic knowledge of programming (Python is preferable). The level achieved with the course “Applied Programming for ME” will be more than sufficient.

After attending this course you should have achieved the following objectives:

  • Know the basics of blockchain technology
  • Know how blockchain can be used to implement cryptocurrency
  • Understand how different cryptography tools (hashing, digital signatures) can be used to implement blockchain
  • Can understand how public blockchains (e.g. Ethereum) can be used to design open systems
  • Can understand how private (permissioned) blockchains can be used to design trustful multiparty systems
  • Knows the basics of different framework to implement blochain-based systems (e.g., Ethereum, Hyperledger)
  • Can understand the importance of blockchain in different real world case studies.

Process Predictive Monitoring (Special Topics in IE - 1)

This course covers the fundamentals of predictive monitoring using business process event logs. The course covers mainly the design and implementation of business process predictive monitoring models built using machine learning techniques. As such, it focuses on topics such as feature extraction and engineering from event logs and encoding of event log information. In the second part, the course also discusses techniques for anomaly detection in event logs.

(Unedited!) recordings of this course lectures in 2020 are available at: Click Here

NOTICE

Recruitment

We are recruiting …

Research Topic

Requirement/Preference

Contacts

  • E-mail
  • Or submit bellow document

Recruitment

Your Name (required)

Your Email (required)

Subject

Your Message

Thank you

Jonghyeon Ko
Jongchan Kim
Soohwan Lee
Jongyub Lee
Kubanychbekov Murat
Marco Comuzzi