Screen Shot 2018-11-13 at 11.00.07 AM


Building the Next Generation Recommendation Engine with a Graph Database 
recorded AUGUST 29, 2018 (45 Minutes)

This is the 2nd episode that will continue TigerGraph's Graph Guru series, a free educational webinar series for developers and data scientists.

An effective recommendation algorithm is critical to increase both consumer experience and vendor revenue. Applications include Netflix movie recommendations, Amazon product recommendations, dating website recommendations, etc. 

So, what's the challenge? The problem is that for a given online user, can the past user traces be explored to recommend something the current user might be interested in?  Most existing recommendation engines in production are built on yesterday's data and computed overnight. The main reason they were built on stale data is that there is no real-time scalable solution that can process fresh data. To address complex and dynamically changing business needs, real-time recommendation is needed.

In this Graph Gurus episode, we use the TigerGraph Developer Edition to build a real-time movie recommendation engine on our GraphStudio visual SDK in 30 minutes. The webinar covers:

  • How to use GraphStudio to model your user data.
  • How to use GraphStudio to load CSV data to the graph database.
  • How to explore the loaded data.
  • How a recommendation engine is built in GSQL using a classic collaborative filtering algorithm. 

Attendees will receive links for the public MovieLens 20M dataset as well as graph queries built for a realtime recommendation engine.


Victor Lee, Director of Product Management, TigerGraph  

Renchu Song, Senior Member of Technical Staff, TigerGraph