/Apparel-Recommendation-System

Recommendation system of apparel using text, brand, color and IMAGE of apparel

Primary LanguageJupyter NotebookMIT LicenseMIT

Apparel Recommendation System

Recommendation system of apparel using text, brand, color and IMAGE of apparel

The data for the project was collected from the Amazon api for women apparel product, because of the large variety of the dataset.

Download the required data from :

https://drive.google.com/drive/folders/1TZPsBO01fk0TzTgCAqSlA_kk5V988svo?usp=sharing

The data has various features :

  1. Title of the product, the description of the production in 10-20 words ( most important feature )
  2. Brand of the product
  3. Color
  4. Type/Category of the product

In this project I've tried out various strategies used for the recommendation.

  1. Bag of Words (BOW)
  2. Term Frequency (TF)
  3. Inverse Document Frequency (IDF)
  4. TF-IDF
  5. Word2Vec
  6. CNN, VGG-16 Model ( for image based recommendation )