Retail product embedding generation for product recommendation
Developed an embedding model for a more natural product recommendation and filtering.
Used a customized derivative of word2vec called item2vec to train from raw shwapno retail invoice data.