In this session we'll build a recommendation engine from scratch while paying particular attention to the modelling choices made along the way.
Our solution will be a hybrid which makes uses of both content based and collaborative filtering and we'll be covering the following topics...
- How do I know what nodes to create?
- How should I be naming relationships?
- Are my relationships too specific? Are they too generic?
- How can I tell if I've got hidden nodes in my model?
- How do I deal with time?
- How do I evolve the model as new requirements come in?
...and any other modelling questions you have!
YOU MAY ALSO LIKE:
Modelling a recommendation engine: A worked example
Mark is a graph advocate and field engineer for Neo Technology, the company behind the Neo4j graph database. As a field engineer, Mark helps customers embrace graph data and Neo4j by building sophisticated solutions to challenging data problems.