Skip to content

Implemented rank-based recommendation system and various collaborative filtering models using Python (NumPy, Pandas, Scikit-learn). Addressed sparsity and cold start problems. Evaluated models using MAE, RMSE, and precision metrics.

Notifications You must be signed in to change notification settings

Shuvob4/Recommendation-System

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Recommendation-System

In our project, we tried to implement a recommendation system using amazon beauty products data, we choose a smaller subset of amazon dataset. Dataset was downloaded from https://cseweb.ucsd.edu/~jmcauley/datasets/amazon_v2/, it's publically available for academic purposes. The datasets are in json format and there are two dataset, review_data and meta_data. Due to large file size i could not upload the file.

I have implemented 5 algorithms to build our Recommendation System based on Amazon Product Review (Beauty Product) Dataset. List of Algorithms:

  1. Demographic Model with weighted function
  2. Item Based Collaborative Filtering (KNNWithMeans)
  3. Model Based Collaborative Filtering (SVD)
  4. User Based Collaborative Filtering (SVD++)
  5. User Based Collaborative Filtering (SVD) And used evaluation metric like recall, precision and RMSE.

I have used surprise from scikit-learn. Need to install it if not installed before. If you are using Anaconda Run Below Command

conda install -c conda-forge scikit-surprise

About

Implemented rank-based recommendation system and various collaborative filtering models using Python (NumPy, Pandas, Scikit-learn). Addressed sparsity and cold start problems. Evaluated models using MAE, RMSE, and precision metrics.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published