Interning at HP’s Security Lab in Bristol, UK, is a return home of sorts for Kiki Lagou, who grew up just outside the city. A high school passion for mathematics combined with an interest in its practical and creative potential led her to study for a B. Eng. in computer science from the University of Southampton. She’ll return there after her internship to begin a master’s degree in the same department.
HP: Kiki, what have you been working on this summer?
I've been working on an overall project to increase the resilience of endpoint devices, which are any devices that communicate with the internet, like PCs or mobile phones. An important aspect of endpoint resilience is the ability to detect attacks. My own project focuses on considering how we might use machine learning to help detect such security intrusions on endpoint devices.
HP: How’s it been going so far?
It’s going well, although I haven’t got to the machine learning part yet. Up to now, my work has been focused on data analysis. I've been studying attacks on endpoint devices using an initial set of data, including data obtained through my own testing in the malware lab that we have here at HP Labs. I'm then taking the data that contains known attacks and using that to develop analysis techniques. When I’m done, I’ll apply machine learning to it to see if I can come up with classifications for different kinds of attacks.
HP: What has been challenging about the work?
It's been harder than I thought it would be to pick out the correct features in the data. It's really interesting work in that you get to dig into the complexities of network traffic. But there’s such a broad range of things that it can represent – so it's a challenge to pick out what’s actually significant. You have to keep trying different combinations of things to see what makes a difference. We’re almost there, though!
HP: Have you been applying concepts you were already familiar with from college, or have you had to learn new ones?
That's a really interesting question because I studied both machine learning and data analysis at university, but coming to the research lab and applying my knowledge to real scenarios is helping me understand concepts at a very different level. One of the biggest things that I've experienced is a massive enrichment in my logical thinking process. I feel like I can think so much more logically through a problem space, and then form a strategy to tackle it, than I could before. I came into the internship feeling like it would mostly be a coding challenge – which it was because I had to build a parser to extract useful information from the network traffic – but the logical thinking skills I’ve gained are immense.
HP: Anything else that has particularly struck you about your internship?
I’ve been able to learn so much from the people working here because everybody's really open about their own research and their own educational and technical backgrounds. It’s been really easy to just approach people and ask questions or work through problems with them. At the same time, I feel like my ideas regarding the project are taken seriously. It’s really cool knowing that my work is helping these researchers with what they're doing.
HP: Is the internship influencing what you might do after your master’s degree?
I think so. This internship is affirming that I want to continue with my focus on machine learning and it's also got me interested in security as well. I think it would be really cool to see if I could combine the two in my future academic work.
