A Voice Assistant for IoT Cybersecurity

The Internet of Things (IoT) is becoming more pervasive in the home, office, hospital, and many other userfacing environments (UFEs) as more devices are networked to improve functionality. However, this explosion of networked devices in UFEs necessitates that security systems become easier to help users remain aware of the security of the devices on their network. Users may not have the skills or the time needed to continuously monitor networks of increasing complexity using common open-source tools. Specifically, they are not likely to fully comprehend the data that those tools present, nor are they likely to have a working knowledge of the tools needed to monitor and protect their IoT-enabled network environments. This paper explores development of a system that uses ambient computing to facilitate network security monitoring and administration. Our system is designed to combine machine-learning–enriched device awareness and dynamic visualization of IoT networks with a natural language query interface enabled by voice assistants to greatly simplify the process of providing awareness of the security state of the network. The voice assistant integrates knowledge of devices on the network to communicate status and concerns in a manner that is easily comprehensible. These capabilities will help to improve the security of UFEs while lowering the associated cognitive load on the users. This paper outlines continued work in progress toward building this capability as well as initial results on the efficacy of the system.