Fatemeh Riahi, PhD candidate
Focusing on outlier detection for structured data while pursuing her PhD in the Computational Logic Laboratory, Fatemeh Riahi has developed a new metric based on statistical modeling.
Improving the ability to identify unusual features of individuals, teams or other subjects from a mass of data, the solution can be used to extrapolate common factors behind the success of actors, sports stars or even movies.
“Looking at the English Premiere League, for example, we could detect the good players from the bad by mining the data. We’ve even applied it to the IMDb movie listing site,” says Riahi, adding that one of the key challenges has been modeling the immense amount of statistical information available.
But her solution isn’t just about charting the tenets of success. It can be applied to all kinds of information to find and isolate unusual features. “This metric hasn’t been used before but we’ve proved it can be applied to any kind of data,” she says.
Putting that approach to the real-world test has been part of Riahi’s internship as a legal compliance analyst at the Vancouver arm of software multinational SAP. “We’re looking into the area of customer compliance. The main challenge is trying to find the best software for this – when new features are needed, it’s up to our team to design them.”
But despite the early successes of her computer science career, Iran-born Riahi had different ideas about her future when she was growing up. “I was very keen on mathematics and I really wanted to study the subject at university. But my father wanted me to have more career opportunities and he convinced me to try computing.” Since switching, she’s never looked back.
After completing her undergraduate degree and moving to Canada for a master’s at Dalhousie University, she looked for the perfect PhD position. “I was already reading the work of Oliver Schulte – he eventually became my SFU supervisor – and I found it very interesting. It seemed like this school would be a great fit for me,” she says, adding she’s enjoyed courses in everything from algorithm design to machine learning at SFU.
And while she imagines ultimately furthering her research career in the U.S., where she believes there are many additional opportunities available in her field, Riahi says she’s glad she accepted her father’s life-changing advice all those years ago.
“Computing science can seem intimidating when you’re a young woman looking at your options. It’s easy to feel you’re not good enough and I certainly felt like that when I was 18. But it’s really not as difficult as it seems to enter this field. And when you start addressing problems and finding your own solutions, it’s a great feeling.”