/RecommendationSystem-Word2vec

Recommendation system for customers in an E-commerce site based on Word2Vec embeddings

Primary LanguageJupyter Notebook

RecommendationSystem-Word2vec

Play around with the interactive app by clicking this link

Goal :

  • To build a recommendation engine that recommends a desired number of products to the consumers on an E-commerce website based on the past purchase behavior of the consumers

Data :

  • The E-Commerce data is taken from the UCI repository and can be found here

Solution :

  • Implemented Word2Vec model to form word embeddings for all the products
  • The model can recommend similar products for a given product and also recommend products to customers by looking at most recent purchases ( can be limited such as last 10 products, 6 months etc.. )