When you have textured hair (i.e. curly, frizzy or frizzy hair), it can be difficult to find the right products for your hair type. Even more so if you're concerned about the ingredients!
So why not build a ingredient-based recommendation tool to find new products?
In this project, we use the data scrapped from the French website labelleboucle.fr which is a website selling hair products with clean and natural composition held by the influencer of the same name.
- Pandas
- Numpy
- Selenium
- Scikit-Learn
- Bokeh
- 01.scraping.py: Web scraping data on labelleboucle.fr with Selenium.
- 02.cleaning.ipynb : Data cleaning
- 03.modeling.ipynd : Preprocessing the data for modelisation with word embedding and dimension reduction technique, and visualizing the items with Bokeh.
- Build a proper web application with Flask
Anne-Gaëlle Sng - annegaelleandco@gmail.com
Project Link: https://github.com/annegaelle-sng/cosmetics_recommendation_engine/>