TechRxiv
IEEE_IoT__preprint_.pdf (721.59 kB)
Download file

Intrusion Detection in Smart Homes: A pretest-posttest study of learning modalities for improved Cyber Situational Awareness

Download (721.59 kB)
preprint
posted on 04.05.2022, 21:33 authored by Christopher D. McDermottChristopher D. McDermott, Mathew Nicho, John P. Isaacs

Smart homes are becoming increasingly more com- plex and difficult to defend. This paper explores how humans can interact with intelligent virtual agents to improve their awareness and detection of threats in smart homes. Mica Endsley’s situa- tional awareness (SA) model was adopted and used to assess how participants perceive device activity, comprehend this in the context of their environment, and project this knowledge to detect if a threat exists. A pretest-posttest study demonstrated a statistically significant improvement in accuracy and efficiency when detecting threats using the conversational agents. In addi- tion, a qualitative approach was also taken to examine usability in relation to a users preferred learning modality. Participants reported being more confident at detecting threats across the three elements of the SA model when using the conversational agents as part of a multi-modal approach. 

History

Email Address of Submitting Author

c.d.mcdermott@rgu.ac.uk

Submitting Author's Institution

Robert Gordon University

Submitting Author's Country

United Kingdom