Document Type
Poster
Publication Date
Fall 2020
Abstract
This project will be a meta-analysis of how the popular social media app known as TikTok takes into account image recognition in their machine learning algorithms through the data it analyzes from its users. It will also see how it identifies and pushes the most beautiful to fame and virality. Though we don’t have access to the actual Tiktok algorithm, we are going to use a very similar dataset known as SCUT-FBP5500. We will analyze how it perpetuates toxic western and eastern beauty standards that are only based on far too simple analyses of what is considered beautiful. We will also use a separate study through a scientific study, which analyzes men and women stimulus in response to beauty. We will lastly use an article, which explores the Chinese app called Alipay, and how it uses beauty filters that perpetuate patriarchal ideals over women. This dataset, study, and article will uncover how human nature and sociology can contribute to how algorithms are truly being fed our want to see idealistic beauty. They will also prove how the belief that the algorithm is inherently bad is false, but that human society around the world needs new establishments of what true beauty is instead. Overall, the goal of this project is to understand these examples of beauty algorithms, how they work, the reason they are used in human society, and how we can improve or discourage use of them in our social media apps.
Recommended Citation
Melonio, Priya, "TikTok’s Non-Inclusive Beauty Algorithm & Why We Should Care" (2020). IPHS 200: Programming Humanity. Paper 22.
https://digital.kenyon.edu/dh_iphs_prog/22
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.