As I have researched evolution for this blog, I have noticed that there is a lot of misinformation about evolution online. This inspired me to write a research paper in which I created a machine-learning architecture that detects misinformation about evolution on social media. I was fortunate enough to be invited to present the paper at the 2024 Carnegie Mellon University IDeaS Conference: Disinformation, Hate Speech and Extremism Online. At the conference, I was the only presenter who was not a graduate student, post doctorate or PhD.
Here is the abstract of my research paper: Billions of people use social media daily on platforms that enable the unfiltered expression of personal beliefs. This has facilitated the proliferation of misinformation with the power to adversely impact societal perceptions of contested ideas. While there has been substantial work towards detecting misinformation on social media, most research focuses on political topics. Few studies have specifically explored scientific misinformation and even fewer misinformation about evolution. In this project, modern machine learning methods are used to detect and classify misinformation about evolution on social media, and a thematic analysis is performed on relevant posts. The architectures developed demonstrated high classification performance, with PCA+SVM achieving an 80% accuracy, recall, and F1 score. Among the categories misinformation about evolution was segmented into, the thematic analysis revealed that the most common category was posts that deny that evolution is valid science, while a smaller category explicitly cites religion to deny evolution.
Evolutionary Anthropology 