 Hi, welcome to our work, a cognitive approach to detect cybersecurity events by Sandeep Nair and Ashwin Kumar Ganeshan from the Accelerated Cognitive Cybersecurity Lab at UMBC. And before we start, a quick introduction. What we're trying to do is to apply cognitive computing approaches to improve cybersecurity. Today, the security industry is heavily dependent on the skill of the analyst on their ability to fuse in their mind disparate pieces of information spread across a temporal window to reason over background knowledge that they're familiar with to decide if something might be an attack, especially in advanced persistent threats. Detecting these attacks early in the cyber-kill chain, ideally left of exploit, is extremely hard. And that is what we are trying to push the boundary on in our project. We start by using the unified cybersecurity ontology that tries to combine disparate sources of information such as CVE and sticks with sensor information gathered from the network. And the UI here shows an amalgamation of this information with all the inferred warnings for administrators. We demonstrate the capability of the system by simulating an attack such as ransomware whose first step is to gain access using an attack like WannaCry. Seen on the screen now is a simulated hacker trying to exploit the SMB vulnerability in the system. The exploit is detected by SNORT and our system begins to populate its knowledge graph with this piece of information. Seen on the UI here are the network alerts, the attacker's IP address, the location of the attacker. In the next step, the attacker tries to download an executable to the victim's machine and tries to gain access and encrypt sensitive files. We can see this in the following screen where the attacker is now trying to download an open SSLEXE file to the victim machine. While the file is being downloaded, we can see in our system a set of system alerts being generated for these various file operations. Once the open SSLE file has been downloaded to the victim's machine, the attacker tries to encrypt sensitive files and we can see immediately an alert being generated about the attack with more details being provided about it. This test is merely an example that demonstrates the capabilities of a system based on knowledge graph reasoning. Thank you.