Cypher-based Graph Pattern Matching in Apache Flink
summary
A Cypher engine runs on Apache Flink by translating MATCH patterns into embeddings and Flink operators (leaf, join-embeddings, expand-embedding using iterative joins, filter) over Gradoop’s property-graph abstraction. The prototype returns a graph collection per match, supports variable-length paths (demo uses up to 10 hops), uses a greedy cost-based planner with label/property statistics, is read-only/analytical (not OLTP), and is open source with benchmarks run on billion-edge graphs and a runnable demo via Gradoop.
AI-written summary · about
chapters (8)
- 0:00Introduction to Graph Query
- 3:35Property Graph Model
- 6:01Cypher Query Syntax
- 9:15Integrating Cypher with Flink
- 14:07Query Evaluation and Operators
- 21:57Query Optimization Plan
- 25:12Live System Demonstration
- 30:43Future Work and Q&A
related talks
17mBringing Cypher to Apache Flink
43mDesigning Cypher - a Graph Query Language
12mBatch Data to Graphs: Integrating Neo4j and Flink via JDBC
24mGraph-Powered Job Matcher: From SQL Tables to Neo4j RAG in Production
49mModern Graph Databases
31mLife of a distributed graph database query
14mA Practitioner's Guide to Graphs
55mLet Me Graph That For You