Graph Analysis over Relational Database. Roberto Franchini - Arcade Analytics
0 likes171 views
This document discusses mapping a relational database to a graph model. It introduces graph concepts and defines vertices, edges, and properties. It describes two strategies for mapping - a naive strategy that directly maps the ER model and a naive aggregate strategy that aggregates many-to-many relationships. The strategies are compared and other alterations like aggregating tables or splitting types are discussed. A DVD rental example database is modeled in both the ER and resulting graph models.
Graph Analysis over Relational Database. Roberto Franchini - Arcade Analytics
- 2. Graph Analysis over Relational Database Roberto Franchini, CTO @robfrankie Gabriele Ponzi, Senior Developer @gponzi
- 3. R.Franchini> whoami CTO (hands on) of Arcade Analytics 20 years of experience OrientDB (Tech lead): JDBC driver, full-text and spatial indexes, docker images, evangelism CELI (Tech lead): NLP (@scale) and semantic search for the market leader in social media monitoring JUG leader, tech speaker
- 4. G.Ponzi> whoami Senior developer at Arcade Analytics MSc in Computer Science Academic and work experience in Databases, Big Data and NoSQL world, distributed data handling, information extraction and visualisation Passionate developer always been in love with cutting edge technologies
- 5. Agenda From relational to graph Intro to graph model ER model and similarities ER to graph To flat or not to flat relations Demo
- 6. Goal: RDBMS as a graph turn upside down your plain old RDBMS explore relations as a graph
- 7. Why? Joins are hard to write Joins are for developers Graph model is just the plain old normalized ER-model Visual navigation is for human beings
- 8. Import into a graph database Write custom process to export from RDBMS and import into a graph Neo4j and OrientDB can import from CSVs OrientDB has Teleporter (live sync only in EE version)
- 9. Import into a graph database Hard to keep in sync work on OLD data Time consuming should wait for sync or import Cost $$$ on maintenance
- 12. Graph G = (V,E)
- 13. Graph: reference model Property Graph Model ● Vertices ○ set of outgoing edges ○ set of incoming edges ○ collection of properties (key-value map) ● Edges ○ outgoing tail vertex ○ incoming head vertex ○ label that denotes the type of relationship between its two vertices ○ collection of properties (key-value map)
- 14. Graph model crash course Model pairwise relations between objects Objects: “Nodes”, “Vertices” Relations: “Edges”, “Arcs”, “Links”
- 15. Nodes (Vertices) Objects in the graph, related by edges Can contain information
- 16. Edges Can be directed or undirected Can store information, usually in the form of an edge “weight”
- 17. Use cases
- 20. Recommendation
- 21. More theory
- 22. Special graphs Regular graph: Each node has the same number of neighbours Complete graph: Every pair of nodes is joined by an edge; every possible edge exists Connected graph: Always possible to move from a to b for any {a,b} Strongly connected graph: Directed path from a to b exists for all {a,b} Bipartite graph: Can separate vertices into two groups such that edges only cross between the groups
- 23. Cycles and trees Cycle: a path that leads you back where you started Some graphs contain cycles, others do not Cycle-free graphs are called “acyclic” Tree: connected graph with no cycles
- 24. Algorithms Shortest Path Dijkstra Centrality Page rank Modularity Clustering, then max degree, min degree, average degree
- 26. RDBMS Connector How to perform a Graph Analysis over a data source with a different data model?
- 27. RDBMS Connector How to perform a Graph Analysis over a data source with a different data model?
- 28. RDBMS Connector: General Workflow
- 29. RDBMS Connector: General Workflow
- 30. RDBMS Connector: General Workflow
- 31. RDBMS Connector: General Workflow
- 32. RDBMS Connector: General Workflow
- 33. Mapping strategies Graph Model allows us to perform several optimisations through the aggregation of some info.
- 34. Mapping strategies Arcade provides two different strategies, each with a specific impact on the mapping between the two domains: naive strategy naive-aggregate strategy
- 35. Mapping: naive strategy Basic approach: E-R Model directly translated into a coherent Graph Model as follows: ● Entity Vertex Type ● Relationship Edge Type Pros: ● Fast ● Few mapping rules needed Cons: ● Model overheads
- 38. Mapping: naive aggregate strategy Basic approach + aggregation: N-N Relationships are aggregated ● Entity Vertex Type ● Relationship Edge Type ● Join Table Aggregator Edge Pros: ● Overhead decreasing ● Resulting graph is less complex, then analysis in simpler! ● Better performances Cons: ● More mapping rules needed
- 39. Mapping: naive aggregate strategy
- 41. Strategies comparison Naive strategy output Naive aggregate strategy output
- 42. Other useful model alterations Aggregation of 1 or more tables into a single Vertex Type
- 43. Other useful model alterations Aggregation of 1 or more tables into a single Vertex Type Single table splitting into multiple Vertex TypesOR
- 44. Other useful model alterations Aggregation of 1 or more tables into a single Vertex Type Single table splitting into multiple Vertex TypesOR Aggregation of 1 or more Relationships into a single Edge Type Single Relationship splitting into multiple Edge TypesOR
- 45. Other useful model alterations Aggregation of 1 or more tables into a single Vertex Type Single table splitting into multiple Vertex TypesOR Aggregation of 1 or more Relationships into a single Edge Type Single Relationship splitting into multiple Edge TypesOR New Vertex or Edge Type definition could be useful too
- 46. Demo use case: DVD Rental E-R Model
- 47. Demo use case: DVD Rental Graph Model (without aggregation)
- 48. Demo use case: DVD Rental Graph Model (without aggregation)
- 49. Demo time
- 50. Thanks for your attention!