The iSensor lab is primarily dedicated to socio-technical research to study trustworthiness and other human behaviors related to cyber security on the web. This socio-technical research approach is based on a social-psychological theory, attribution theory, to model a sensor system, a conceptual framework of trustworthiness attribution which focuses on understanding how anomalous behavior of an individual can be detected by observers. We study how humans attribute meaning of the behavior of others with limited resources, i.e., in text-only, cue-lean environments.
One of the applications of this research is to the study of ‘insider threat,’ which refers to situations where a critical member (or members) of an organization behaves against the interests of the organization, in an illegal and/or unethical manner. Studies adopt the concept of human-observed changes in behavior as being analogous to a group of “sensors” on a computer network. Using online team-based game-playing, this study seeks to re-create realistic insider threat situations in which human sensors have the opportunity to observe changes in the behavior of a focal individual. A full-scale experiment has been designed and conducted for data collection and analysis.This predictive approach formulates a novel understanding of insider threats, providing contemporary ways of understanding anomalous behavior in virtual space, and thus contributing to cyber infrastructure security.
Several experiments were conducted using online team-based game-playing. This study endeavored to re-create realistic situations in which human sensors have the opportunity to observe changes in the behavior of a focal individual. Transcripts of communications were examined to recognize how human sensors attribute meaning. Results of this study show that observed changes in behavior can identify a downward shift in the trustworthiness of a critical member. Human sensors triggered a two-stage recursive attribution mechanism to make sense of suspicious behavior. The contributions of this socio-technical study lie in its capability to tackle complex insider threat problems by adopting a social psychological theory on predicting human trustworthiness in a virtual collaborative environment. The two-stage recursive attribution mechanism contributes to the behavioral anomaly detection computational modeling of intelligence sensor in virtual worlds.
Using online games to simulate threat situations helps to generate rich data that can offer insights into complex security problems. This predictive approach formulates a novel understanding of insider threats, providing contemporary ways of understanding anomalous behavior in virtual space, and thus contributing to cyber infrastructure security.
Another modeling strategy for an intelligence sensor system, or i-Sensor, which can comprehend human’s virtual dialogues in Computer Mediated Communications (CMC), focuses on processing those dialogues to understand trustworthiness detected among humans in a virtual collaborative group.
Future research: Continuing the effort of building education and research capacity in ubiquitous secure technology. iSensor Lab purposefully seeks to utilize existing resources, strengths, and the efforts of the University’s members, students, and research participants, to pursue socio-technical research with rigor and relevance, and to develop a simulated virtual lab to study human interaction during crisis management in the virtual world. This initial development effort will include a set of crisis and emergency scenarios that guide virtual participants to collaborate and coordinate as virtual teams to respond in disaster situations. Interaction will be monitored and studied to understand the patterns of coordinated relationships that emerge during crisis. Virtual participants will include scientists, government agencies (e.g., FEMA, DHS), intelligence community, members of emergency response teams in physical context, as well as geographically dispersed corporations and organizations.
iSensor Research is a project owned by