In this project, I processed ingredient lists for 1472 cosmetics on Sephora via word embedding, then visualized ingredient similarity using a machine learning method called t-SNE and an interactive visualization library called Bokeh.
These days, because I wear a mask every day, skin problems are frequently occurring. For this, I want to try new cosmetics, but sometimes new attempts can worsen skin troubles. There's ingredient information behind cosmetics, but it's difficult to understand and what to avoid.
To help with this, I created a content-based recommendation system in which content becomes a chemical component of cosmetics.
This project is licensed under the MIT License