Gated recurrent unit-based expiration time prediction for indicators of compromise in cyber threat intelligence
The Computer Journal · 06.08.2026
DOI: 10.1093/comjnl/bxag078 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Computer Journal, online erschienen am 06.08.2026. Verfasst von Yuan-Cheng Lai und 2 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben.
Abstract (Originalsprache)
Cyber Threat Intelligence (CTI), which encompasses information about attackers, is required to effectively counter these threats. Indicators of Compromise (IoC) are fundamental elements of CTI, and system security can be ensured by checking the IoCs. IoCs, however, have a limited lifetime because attackers may discard this attack. Invalid IoCs are persistently stored, resulting in unnecessary resource waste and an undeserved block. Therefore, this paper investigates how to automatically remove expired IoCs by using Gated Recurrent Unit (GRU)-based expiration time prediction (GETP), which consists of seen time estimation and expiration time estimation. The former predicts the next seen time of each IoC, whereas the latter adds an appropriate tolerance as its expiration time. There are two approaches for determining the tolerance length: Direct Prediction and n-Prediction. With the GRU model, Direct Prediction directly predicts tolerance, whereas n-Prediction predicts the multiple of the variance of the seen time. The experimental results show that GETP outperforms traditional exponential smoothing and Long Short-Term Memory in predicting seen time accuracy by at least 20.4% and 15.0%, respectively. For the expiration time estimation, Direct Prediction and n-Prediction improve the previous method, Decay Function, by 30.2%–46.4% and 37.5%–50.5%, respectively, on the IoC fault rate.
Zitieren
APA
Lai, Y., Lin, C., & Wang, P. (2026). Gated recurrent unit-based expiration time prediction for indicators of compromise in cyber threat intelligence. The Computer Journal. https://doi.org/10.1093/comjnl/bxag078
Vancouver
Lai Y, Lin C, Wang P. Gated recurrent unit-based expiration time prediction for indicators of compromise in cyber threat intelligence. The Computer Journal. 2026. doi:10.1093/comjnl/bxag078
BibTeX
@article{Lai2026,
author = {Lai, Yuan-Cheng and Lin, Chih-Chung and Wang, Pei-Ru},
title = {Gated recurrent unit-based expiration time prediction for indicators of compromise in cyber threat intelligence},
journal = {The Computer Journal},
year = {2026},
doi = {10.1093/comjnl/bxag078}
}
EUVIMED