EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Which Delay Matters? A Cause-Aware Waiting-Time Decomposition for Internal-Control Risk Detection in Business Process Event Logs

Yu-Ting Lin, Chia-Hao Huang, Wei-Jen Chen*

SOCIAL AND MANAGEMENT SCIENCES RESEARCH JOURNAL · 20.08.2026 · 4 (2) · 59-75

DOI: 10.66238/smsrj.v4i2.69 ↗

Kurz gefasst

Fachartikel in der Zeitschrift SOCIAL AND MANAGEMENT SCIENCES RESEARCH JOURNAL, online erschienen am 20.08.2026. Verfasst von Yu-Ting Lin und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Operational delay and internal-control risk are normally analysed by different functions with different tools, and the assumption that accelerating a process also improves control is rarely tested on real data. This paper asks whether process delay carries information about internal-control risk and, if so, which component of delay carries it. We introduce a pool-level, capacity-aware decomposition of pre-activity waiting time into resource contention, calendar unavailability and residual delay that is exactly computable on the kind of log real finance functions actually own: one timestamp per event and anonymised resources. The contention window of a waiting interval is obtained in closed form from a k-th-largest departure rule, giving an O(n log k) algorithm. We combine the decomposition with a prefix-safe observation design in which every item is observed at its first control point, so that all features precede the control decision that defines the label. On three public event logs (two BPI Challenge 2020 expense-approval logs and the CoSeLoG receipt-phase log; 18,497 analysed items) we find that waiting is dominated by calendar unavailability (75.3-76.1%), that total waiting is nevertheless a robust item-level risk indicator (odds ratio per standard deviation 1.27, 1.09 and 1.46 after quarter, role, amount and off-hours controls), and that ranking items by waiting time alone retrieves 21.3%, 17.3% and 24.2% of all control exceptions within a 10% inspection budget. Conditional on total waiting, the working-time component carries the risk in two of the three logs while calendar-idle waiting does not. Two negative results qualify the practice: the incremental AUC of delay features over context and conformance features is small (+0.011 to +0.015), and the delay-risk association exists at item level but disappears at period level once secular trend is removed. Delay is therefore a cheap and reliable signal for audit sample selection, not a period-level control indicator.

Zitieren

APA

Lin, Y., Huang, C., & Chen*, W. (2026). Which Delay Matters? A Cause-Aware Waiting-Time Decomposition for Internal-Control Risk Detection in Business Process Event Logs. SOCIAL AND MANAGEMENT SCIENCES RESEARCH JOURNAL, 4(2), 59-75. https://doi.org/10.66238/smsrj.v4i2.69

Vancouver

Lin Y, Huang C, Chen* W. Which Delay Matters? A Cause-Aware Waiting-Time Decomposition for Internal-Control Risk Detection in Business Process Event Logs. SOCIAL AND MANAGEMENT SCIENCES RESEARCH JOURNAL. 2026;4(2):59-75. doi:10.66238/smsrj.v4i2.69

BibTeX
@article{Lin2026,
  author = {Lin, Yu-Ting and Huang, Chia-Hao and Chen*, Wei-Jen},
  title = {Which Delay Matters? A Cause-Aware Waiting-Time Decomposition for Internal-Control Risk Detection in Business Process Event Logs},
  journal = {SOCIAL AND MANAGEMENT SCIENCES RESEARCH JOURNAL},
  year = {2026},
  volume = {4},
  number = {2},
  pages = {59--75},
  doi = {10.66238/smsrj.v4i2.69}
}
Kostenlos starten →