Grant: Discovery Grants Program - Individual

While people use the convenient and fast online resources to collect information daily, these systems may contain incorrect or biased content that has the potential to cause harm, especially if we believe what we read without further research or professional advice. During the last several decades, the spread of harmful online content has raised concerns by both the scientific community and the general public, often due to algorithmic bias issues and varying opinions of different online communities. The overall reaching goal of this research program is to analyze the effect of online harmful content on people's attitudes. In particular, the research will focus on providing insights to better evaluate and design social media and search system to mitigate online harm. With a computer science background, this research will solve these problems from a computer science perspective. This research aligns with the Government of Canada's commitment to take meaningful actions to combat hate speech and harmful content online. The specific objectives of my proposed research program are as follows: i) Study online search behaviour in the presence of harmful content returned by the search systems. This will provide insights into the user behaviour to predict and better evaluate information retrieval systems. ii) Explore potential implications of the spread of harmful content in social media platforms. The findings will provide design recommendations to neutralize and mitigate the effect of online hate speech, and to moderate, monitor, and achieve platform governance. This research lines up with what The Canadian Liberal government recently proposed in introducing new regulatory frameworks to address harmful content through social media platforms and online communications. iii) Focus on the problem of harmful content within a specific age group. Many prior studies looked at general users' susceptibility to misinformation. Given that we lose peripheral acquaintances with age, older adults may likely have stronger relationships with people they follow on social medial. With a small social circle and low digital literacy, online connections might have a huge impact on older adults' views. Thus, it is important to understand how older adults experience social media with the presence of misinformation. These findings will provide new design recommendations to consider when building online platforms to address older adults' special needs. As a whole, this work will lead to improving current systems to be fairer, more accessible, and generally less biased. While each project targets a different platform or user demographics, together they will contribute to building systems that better support people's decision making by mitigating online harm. The proposed research program will further provide ample opportunities for students to gain experience in multi-disciplinary areas such as modeling and prediction of human behavior as well as large scale data processing.