Recommendation System for Amazon Alexa E-Commerce Application

• Researched, planned and developed a personalized product recommendation engine from scratch, to be deployed as a micro service for ecommerce shopping cart applications.
• Did detailed research, including studying research papers and evaluated types of recommender systems.
• Trained, tested and developed a production ready recommender system using Tensorflow, sklearn, numPy, pandas, sciPy, Flask, Flask-PyMongo and MongoDB (NoSQL). It was built on Cosine similarities between TF-IDF vectors in vector space representation algorithm for content-based filtering combined with Matrix Factorization model using WALS algorithm to optimize the loss function for collaborative filtering.
• Collaborated with the apigee and Amazon Alexa teams to integrate the output of the recommender system to an Amazon Echo device.

This repository consists of a Recommendation engine using the Collaborative filtering and Content Based filtering methods in TensorFlow.

Dataset used: Amazon ecommerce sales dataset of ~10000 training examples, with ~15 features each was split for use in training and testing of the algorithms.

1. collab .py contains code for the collaborative filtering aspect of the recommendation engine.
2. contentTensor .py contains code for the content-based aproach of the recommendation engine.
3. flaskTest1 .py contains code for exposing both content-based and collaborative filtering methods as an API that returns JSON output using Python Flask.