notes.txt

Recommendation system model that I am trying to implement
- Recommends list of products from respective skincare products descriptions (skin concerns) 
  Lets say user A used an item that coherent with a unique skin type and its skin concerns 
  Therefore, the system will recommend top related products with these features

Procedures
1. Web scrape data from sephora malaysia using selenium 
   Among the data attributes scraped are product brand, name, skin_type, skin_concern and rating

2. Data cleaning and preprocessing
   - Drop duplicate data and data with missing field
   - Remove RM to convert price to float

Requirement
1. Anaconda (Python 3)
2. Jupyter Notebook - editor 
3. Selenium & Geckodriver - driver to access web browser for web crawl data 
4. Pandas 
5. Regex - to store regular expressions like (%^>) into dataframe
6. Sklearn - tfidfvectorizer, cosine similarity