Recommending products, whether we are talking about movies, gardening tools, or meetups, share some common patterns. Here are some common-sense assertions that can lead to a good recommendation:
- Products in the same categories to a product already bought are more likely to be useful to the user. For instance, if you buy a rake, it probably means you like gardening, so a lawnmower could be of interest to you.
- There are some products that often get bought together, for instance, printers, ink, and paper. If you buy a printer, it is natural to recommend the ink and paper other users also bought.
We are going to see the implementations of those two approaches using Cypher. We will again use the GitHub graph as a playground. The important parts of its structure are shown in the next schema:
It contains the following entities:
- Node labels: User, Repository, Language, and Document
- Relationships:
- A User node owns or contributes to one or several Repository...
