Recommender Systems Methods and Usage of Graphs for Recommendations


Recommender systems have been around for a long time now and have proven to be effective in:
• solving the information overload problem
• increasing user satisfaction
• increase in sales due to cross selling
• increase in user engagement etc.
Recommender systems are a type of information filtering systems which can be critical for some types of businesses such as e-commerce, streaming services etc. as they can drive huge revenue uplift.

There are multiple traditional machine learning algorithms used across industries to build recommendation engines such as Content -based recommendations, collaborative filtering based, hybrid approach etc. These models pose certain challenges such as lack of diversity in recommendations, cold-start problem, sparsity problems etc.

There has been a surge in usage of graph-based models to generate recommendations. These include PinSage, GraphSage, GNNs, Bayesian graph-based approaches, random walk etc. The choice of best fit approach depends on various aspects such as complexity of problem, expected output, type of data and size, time constraints etc.

This tutorial aims to walk the participants through a brief introduction of recommender systems, explanation of traditional algorithms along with their limitations. Next, we will discuss how graphs are proving to be an effective resource in developing effective recommender systems. Additionally, we will also discuss other challenges related to recommender systems including identifying the method best suited to measure effectiveness of a recommendation system in industrial set up and explain ability of recommendations. We discuss scenarios where it can be essential for models to support expandability capabilities and how can that be achieved in recommender systems.

Background Knowledge:

Python, Jupyter Notebook


Harshita is an Applied Scientist at Amazon based out of Seattle. She creates ML models that help various teams across AWS Sales and Marketing drive important businesses decisions. She specializes in generating actionable insights from data and quantifying business impact of various initiatives. Harshita is a ML enthusiast who graduated from Syracuse University with a Masters in Data Science in 2021.

Open Data Science




Open Data Science
One Broadway
Cambridge, MA 02142

Privacy Settings
We use cookies to enhance your experience while using our website. If you are using our Services via a browser you can restrict, block or remove cookies through your web browser settings. We also use content and scripts from third parties that may use tracking technologies. You can selectively provide your consent below to allow such third party embeds. For complete information about the cookies we use, data we collect and how we process them, please check our Privacy Policy
Consent to display content from - Youtube
Consent to display content from - Vimeo
Google Maps
Consent to display content from - Google