A product based recommendation system which uses Machine learning algorithm such as KNN and cosine similarity and also uses MongoDB as a database which stores the user data for a semi-collaborative filtering.

Accuracy :

Calculated accuracy using nDCG.

Some randomly selected product efficiency:

  1. Batman killer croc takedown figures: nDCG=0.917

  2. Star Wars Movie Heroes Yoda: nDCG=0.942

  3. Harry Potter Hogwarts Bookmarks: nDCG=0.9406

Technology Used in this project:

  1. Pandas
  2. Numpy
  3. Sklearn
  4. MongoDB as Databases
  5. Streamlit for UI



Home UI:

Database structure:

Result UI:


View Github