Human Evaluation and Tool Development for Online Content Credibility

Thesis event information

Date and time of the thesis defence

Place of the thesis defence

L4, Linnanmaa campus

Topic of the dissertation

Human Evaluation and Tool Development for Online Content Credibility

Doctoral candidate

Master of Engineering Junhao Li

Faculty and unit

University of Oulu Graduate School, Faculty of Information Technology and Electrical Engineering, Center for Ubiquitous Computing

Subject of study

Computer Science

Opponent

Associate Professor Evangelos Karapanos, Cyprus University of Technology

Custos

Professor Simo Hosio, University of Oulu

Add event to calendar

Human Evaluation and Tool Development for Online Content Credibility

The Internet has become a primary source of information for people worldwide. With their heavy reliance on user-generated content, social media platforms play an important role in this change. People are making decisions based on online sources, such as social media. However, the credibility of user-generated content is being increasingly questioned, ranging from verified truths to unfounded rumours. The prevalence of misleading health information is one of the most worrying issues. Misleading health information can lead to harmful behaviours. Although measures such as content moderation are taken to combat misinformation, the scale of this problem requires sophisticated solutions. This underscores the urgent need for research to be conducted on the credibility of online content and solutions to manage misinformation.

This thesis is conducted with two core objectives: 1) To gain an understanding of the credibility of social media content from the perspectives of laypeople. 2) To investigate potential solutions to help individuals adapt to the current online environment flooded with misinformation.

This thesis employed a combination of methods to examine factors that affect people's credibility assessment process based on empirical investigations. The studies were designed to capture the behaviour of real people and included controlled online experiments to explore how participants evaluate the credibility of different types of information. In addition, the research investigated the use of advanced AI tools to simplify complex medical content. This approach aimed to enhance users’ understanding and improve their ability to make informed decisions.

The findings of this thesis revealed laypeople's perspectives on the challenges of social media credibility. The research identified multiple important factors affecting the credibility assessment process. Furthermore, the research on using LLMs to simplify complex medical content demonstrated promise in improving accessibility to complex medical literature.

As the thesis's final contribution, the findings highlighted the complex challenges and trade-offs in combatting misinformation. It provided promising solutions for combatting misinformation and improving the accessibility of complex medical literature. Furthermore, it offered valuable advice for researchers who aim to combat misinformation.
Last updated: 2.6.2025