Building a Recommendation Engine with Neo4j and Python
summary
Builds a real meetup.com recommendation engine using Neo4j to glue data, Cypher for graph queries and Python for analysis and pre/post processing; imports use load CSV with periodic commit for large member/event sets, computes topic similarity and runs community detection (walktrap) in igraph (279 topics reduced to 81 clusters), writes cluster nodes back into Neo4j, and combines content and collaborative signals to rank groups and upcoming events with simple weighting of member-affiliation versus similarity.
AI-written summary · about
chapters (11)
- 0:00Recommendation engines intro
- 2:29Generic recommendations fail
- 6:18Neo4j property graph models
- 13:48Load Meetup data into Neo4j
- 19:15Indexing and constraints in Neo4j
- 28:13Find similar groups with Cypher
- 37:12Load member data and relationships
- 45:10Query member interests and clusters
- 54:06Recommend events to users
- 1:02:28Combine group and topic recommendations
- 1:08:59Scaling with Neo4j
related talks
01h 38mBuilding a recommendation engine with Python and Neo4j
51mBuilding a Meetup Recommendation Engine with Spring Data Neo4j
46mYour own recomendation Engine with Neo4j AND Reco4PHP
38mRecommendation Engine with Neo4j, Piece of Cake
21mUnlocking Real-Time News Recommendations: The Power of Graph Relationships in Neo4j
27mUsing Graph Theory to Build a Recommendation Engine in Node js
19mScalable Graph Processing with Parallel Graph AnalytiX
55mBuilding a Recommendation Engine with Machine Learning Techniques