In today’s connected world, information of every hue––intellectual property, health records, personal communications, government records, financial information––is accessible over the Internet and other networks. The nearly ubiquitous and on-demand access to a wide variety of information has given rise to an array of services that improve individual and organizational efficiencies, and our overall quality of life. At the same time, such open access poses substantial risk to individuals and organizations, as has been demonstrated by recent high-profile hacking events.
The University of Southern California’s Information Sciences Institute (USC/ISI) is sponsored by IARPA under the CAUSE program to research technologies that will counter cyber threats by anticipating them before they manifest in the form of an actual attack. The EFFECT research team, led by USC/ISI, includes a team of experts from Arizona State University, Raytheon BBN, Hyperion Gray, Lockheed Martin ATL, Ruhr University Bochum. The team is exploring methods to forecast cyberattacks by integrating information from a variety of sensors within a robust machine inference framework.
To find traces of early planning activity by malicious actors, researchers will collect data from unconventional sources, that include dark web and social media sites, forum discussions and others, develop methods to deal with large volumes of heterogeneous information that include unstructured natural language text, structured data, and public network traffic data, and analyze these data streams to generate robust warnings of pending cyberattacks.
The age-old saying to be forewarned is to be forearmed rings particularly true in the case of cyber threats. For society to continue enjoying the benefits of an open, worldwide Internet, it is critical that we tame the rapidly growing cyber threats posed by a variety of state and non-state actors.
![DarkEmbed: Exploit Prediction with Neural Language Models [link]](http://bibbase.org/img/filetypes/link.svg "Link")
![DarkEmbed: Exploit Prediction with Neural Language Models [pdf]](http://bibbase.org/img/filetypes/pdf.svg "Paper")