Reza Shafiloo aims to make recommendation systems more transparent and less biased

Various AI-based recommendation systems significantly influence our lives, guiding us to new music, determining what we watch on streaming services, and even influencing our online shopping choices. While these systems simplify our lives in many ways, their use is not without problems.

Sometimes, these suggestion systems might have a problem. They might not be fair in recommending things to everyone. I'm working on making these systems fairer so that everyone gets a fair chance to see good suggestions. Also, I'm trying to make sure these systems explain why they suggest something. It's like knowing the reason behind a recommendation, and it helps in making sure the suggestions are fair and unbiased.

Reza Shafiloo

Addressing these issues is Reza Shafiloo, a doctoral researcher from Iran, who studies the fairness and transparency of recommendation systems at Tampere University. Shafiloo stumbled upon the field of recommendation systems by chance, attending various seminars at the start of his Master's studies. “I heard about recommendation systems and their bias issues, and I got really interested in understanding and solving these problems,” he explains.

Photo: Antti Yrjönen

Shafiloo's research bridges technology and human experience. “As I learned more about recommendation algorithms, I saw how they affect people's experiences and our society. I'm excited about making these systems fairer and more transparent.”

Shafiloo believes that in the future, human-machine interaction will become more intuitive and natural. 

I think a future where machines are present in various aspects of our daily lives, becoming indispensable problem-solvers. As technology continues to advance, human-machine interaction will likely evolve to be more intuitive and natural. We may see increased collaboration between humans and machines in fields such as healthcare, education, and beyond, enhancing our ability to solve complex problems.

Reza Shafiloo

The multidisciplinary approach of Tampere University's CONVERGENCE project attracted Shafiloo to apply for the programme. He believes that collaboration between different disciplines helps view research challenges from new angles. “Cross-disciplinary collaboration is invaluable as it can uncover new insights, offer innovative solutions, and challenge assumptions that might not have been apparent within a single discipline,” Shafiloo notes.

Shafiloo's research has the potential to impact how we interact with online shopping sites, search engines, and music services in the future. Currently, a significant issue is the lack of transparency, as often even the developers of recommendation algorithms do not fully understand the basis of the suggestions presented to users. “By improving the fairness and transparency of recommendation systems, we can ensure that everyone receives unbiased recommendations,” Shafiloo describes.

Moving to Finland felt like a natural choice for Shafiloo. “Finland is a peaceful and safe country for students. The conditions are perfect for focused academic work.” He admits to having concerns about Finland's cold climate but adds, “Now that I've experienced it, I find the weather quite tolerable.”

Photo: Antti Yrjönen