Dynamic human resource selection for business process exceptions

• DOI: http://doi.org/10.1002/kpm.1591
• Date: 01 January 2019
• Author: Jinwoo Kim,Jooseok Lee,Seunghoon Lee,Injun Choi
• Published date: 01 January 2019

RESEARCH ARTICLE

Dynamic human resource selection for business process

exceptions

Jooseok Lee |Seunghoon Lee |Jinwoo Kim |Injun Choi

Department of Industrial and Management

Engineering, Pohang University of Science and

Technology (POSTECH), Pohang, South Korea

Correspondence

Injun Choi, Department of Industrial and

Management Engineering, Pohang University

of Science and Technology (POSTECH), Nam‐

gu, Pohang 37673, South Korea.

Email: injun@postech.ac.kr

Funding information

Korea Small and Medium Business Adminis-

tration, Grant/Award Number: 10031496;

National Research Foundation of Korea(NRF),

Grant/Award Number:

NRF‐2017R1A2B4008029

A key capability of today's organizations is to flexibly and effectively react to unex-

pected events. A critical case of an unexpected event is sudden unavailability of

human resources, which was not properly addressed by existing resource allocation

approaches. This paper proposes a systematic approach that analyzes event logs to

select suitable substitutes if the initial human resources become unavailable. The

approach uses process mining and social network analysis to derive a metric called

degree of substitution, which measures how much the work experiences of the

human resources overlap, from the two perspectives: task execution and transfer of

work. Along with the metric, suitable substitutes are also identified. A simulation dem-

onstrates that the approach identifies suitable substitutes more effectively and accu-

rately than existing allocation methods such as role‐based allocation or random

allocation. The proposed approach will increase the effectiveness of dynamic alloca-

tion of human resources, especially in an exceptional situation.

1|INTRODUCTION

During execution of business processes, numerous exceptions can

occur. For example, core human resources can become unavailable.

Today's organizations must react appropriately to these situations to

ensure that their processes are executed successfully (Dumas, La

Rosa, Mendling, & Reijers, 2013; Kim, Choi, & Park, 2011; Kim, Lee,

Lee, & Choi, 2017; Reichert & Weber, 2012). Responses to such

exceptions include trying alternative paths, adding supplementary

behaviors, and changing the resource allocation (Reichert & Weber,

2012). The last type of response is especially important when excep-

tions occur in business processes in which human performers have

important roles. It includes delegation, escalation, deallocation, and

reallocation of work items (Russell, van der Aalst, Hofstede, &

Edmond, 2005).

Many business process management systems, however, use

primitive approaches such as the “push or pull”mechanism to dis-

tribute work items to performers (Huang, van der Aalst, Lu, &

Duan, 2010; Liu, Cheng, & Ni, 2012; Liu, Wang, Yang, & Sun,

2008). Thus, work items are often assigned to “too few, too many,

or even the wrong set of performers”(Kumar, van der Aalst, &

Verbeek, 2002). A critical case is that a work item is assigned to

a performer who is absent or overloaded. Some research has

proposed advanced human‐resource allocation mechanisms to avoid

these errors. However, the mechanisms do not consider exceptional

situations such as sudden unavailability of the performers. Such sit-

uations cannot all be anticipated and incorporated in the allocation

mechanism in advance, so appropriate substitutes must be identi-

fied quickly. That is, alternative performers who can minimize the

impact of the exceptions must be identified; this task has not been

properly addressed by previous studies.

Furthermore, existing studies do not consider the perspective of

transfer of work, which can significantly affect the efficiency of busi-

ness processes (Kumar, Dijkman, & Song, 2013). Transfer of work,

refers to passing work items from one task to the next, often contrib-

utes a significant portion of delays in execution of business processes

(Dustdar & Hoffmann, 2007; Hearn & Choi, 2013; Lee, Sung, Song, &

Choi, 2015). Accordingly, both task execution and transfer of work

must be considered.

This paper proposes a systematic approach to rapidly identify

suitable substitutes by considering task execution and transfer of

work. Specifically, the approach uses a metric called degree of substi-

tution (DoS) to quantify the overlap in the work experiences of human

resources. DoS is derived using process mining techniques and social

network. The approach also constructs a substitution network to visu-

alize the DoS. A simulation demonstrates that the approach identifies

Received: 3 April 2017 Accepted: 15 November 2018

DOI: 10.1002/kpm.1591

Knowl Process Manag. 2019;26:23–31. © 2018 John Wiley & Sons, Ltd.wileyonlinelibrary.com/journal/kpm 23