Article: Human and machine partnership: natural language processing of army insider threat hub data Journal: International Journal of Applied Decision Sciences (IJADS) 2025 Vol.18 No.7 pp.1 - 22 Abstract: Threats to organisational efficacy and wellness may come from competitors (external threat) or from trusted agents (insider threat). Countering the insider threat is an imperative for the security of governments, the military, businesses, and all other organisations and institutions that employ people. This paper presents a case prioritisation system that utilises a deep learning classification model trained on expert evaluated insider threat cases to label cases as 'negligible', 'low', 'medium', or 'high' threat level. This classification model enables a partnership between machine and human that focuses human effort for the greatest impact. To evaluate the models created, the authors created a metric called 'detection accuracy rate' that measured correct prediction and over-estimations of threat, with the best model achieving a 96% detection accuracy rate. Inderscience Publishers - linking academia, business and industry through research

Title: Human and machine partnership: natural language processing of army insider threat hub data

Authors: Saleem Ali; Hayden Deverill; Joseph Lindquist; Jonathan Roginski

Addresses: School of Engineering, Brown University, Providence, RI 02912, USA ' Department of Mathematical Sciences, United States Military Academy, West Point, NY 10996, USA ' Department of Mathematical Sciences, United States Military Academy, West Point, NY 10996, USA ' Department of Mathematical Sciences, United States Military Academy, West Point, NY 10996, USA

Abstract: Threats to organisational efficacy and wellness may come from competitors (external threat) or from trusted agents (insider threat). Countering the insider threat is an imperative for the security of governments, the military, businesses, and all other organisations and institutions that employ people. This paper presents a case prioritisation system that utilises a deep learning classification model trained on expert evaluated insider threat cases to label cases as 'negligible', 'low', 'medium', or 'high' threat level. This classification model enables a partnership between machine and human that focuses human effort for the greatest impact. To evaluate the models created, the authors created a metric called 'detection accuracy rate' that measured correct prediction and over-estimations of threat, with the best model achieving a 96% detection accuracy rate.

Keywords: insider threat; natural language processing; NLP; classification; machine learning.

DOI: 10.1504/IJADS.2025.146569

International Journal of Applied Decision Sciences, 2025 Vol.18 No.7, pp.1 - 22

Received: 26 Sep 2024
Accepted: 07 Apr 2025
Published online: 04 Jun 2025 *

Title: Human and machine partnership: natural language processing of army insider threat hub data

Keep up-to-date