End-users' online behavior can significantly affect the reliability and security of next-generation software systems. For instance, skipping repeated requests to update software or ignoring security warnings while visiting unknown websites, while extremely dangerous, are not uncommon. Although end-users' actions (or inactions) often open up the opportunity for cyber-attacks, the lack of emotional appeals and poor design of the current software update/warning messages are to blame to a large extent for such risky behavior, which is addressed as follows. First, this project identifies the limitations of the current software update/warning messages from an affective-cognitive perspective, and addresses the shortcomings by incorporating emotional appeals that command attention and evoke emotions more effectively. Also, message designs are changed adaptively based on the context and user's past behavior to counter the habituation-effect. Finally, to increase the compliance rate, this project develops emotional education and inoculation strategies that enable users to recognize and avoid the temptation associated with risky online behavior. Successful completion of this project will greatly advance the understanding of the impact of emotion on the decision making process in regards to safe online behavior, and will revolutionize the way software vendors communicate the risks of running vulnerable software and warn users. The research findings will be published in premier academic venues, and will be disseminated through project website. Research agendas will be integrated with the curriculum development activities. Existing programs at UConn will be leveraged to promote diversity and engage students from underrepresented communities.
