Summary

We learned how to build an explicit feedback type recommendation system. We implemented a predictions model with the Spark MLlib collaborative filtering algorithm that learns from past sales data and makes ratings-based recommendations about products to customers. The algorithm, as we have come to know, made its tailored product predictions on unknown customer-product interactions. 

We used Spark's support for recommendations to build a prediction model that generated recommendations for unknown customer-product interactions in terms of sales leads and past weapons sales data. We leveraged Spark's alternating least squares algorithm to implement our collaborative filtering recommendation system.