https://towardsdatascience.com/how-to-build-a-collaborative-filtering-model-for-personalized-recommendations-using-tensorflow-and-b9a77dc1320

In this article, I will step you through how to use TensorFlow’s Estimator API to build a WALS collaborative filtering model for product recommendations. Recently, my colleague Lukman Ramsey published a series of solutions detailing how to build a recommendation model — read those solutions for context on what recommendations are and how to set up an end-to-end system.

In this article, I’m replacing the use of Pandas in the original solution by Apache Beam — this will permit the solution to scale to larger datasets easier. Because the context exists in the solution, I’ll simply dive into the technical details here. The complete source code is on GitHub.