3. ML/AI & Graphs Synergy
#1 Graph Platform Neo4j Customer Use Cases Unite Neo4j
Neo4j SuperiorityGraph DB Popularity
4.
5. 7/10
20/25
7/10
Top Retail Firms
Top Financial Firms
Top Software Vendors
Anyway You Like It
Creator of the Property
Graph and Cypher language
at the core of the GQL ISO
project
Thousands of Customers
World-Wide
HQ in Silicon Valley, offices
include London, Munich,
Paris & Malmo
Industry Leaders use Neo4j
On-Prem
DB-as-a-Service
In the Cloud
16. Rise
of(Graph)
Data
Science
- Get more insight out
of the data you
already have.
- Add relationships
into your existing
workflows.
- Make better, more
accurate predictions.
Why Do Graph Data Science?
17. Neo4j Graph Data
Science Library
Scalable Graph
Algorithms & Analytics
Workspace
Native Graph
Creation & Persistence
Neo4j
Database
Visual Graph
Exploration
& Prototyping
Neo4j
Bloom
Practical Integrated Intuitive
Rise
of(Graph)
Data
Science
Neo4j for Graph Data Science
18. 2014 2015 2016 2017 2018 2019 20202013
The Rise of the Graph Database
2010 - 2020: From Model to Category
Full
Potential of
the Property
Graph Model
19. Oct
2020
1.
2.
3.
Rank
4.
5.
6.
7.
8.
9.
10.
Neo4j
Microsoft Azure Cosmos DB
ArangoDB
OrientDB
Virtuoso
Amazon Neptune
JanusGraph
GraphDB
FaunaDB
Dgraph
DBMS Database Model
Graph
Multi-model
Multi-model
Multi-model
Multi-model
Multi-model
Graph
Multi-model
Multi-model
Graph
Score
Oct
2020
Sep
2019
51.33
32.01
5.55
5.47
2.57
2.48
2.40
2.09
1.79
1.68
+0.71
+0.34
-0.25
-0.01
+0.01
+0.13
+0.06
+0.67
+0.07
+0.06
Full
Potential of
the Property
Graph Model
The Rise of the Graph Database
2010 - 2020: From Model to Category
20. Highest possible scores in:
● Performance
● Scalability
● Workloads
● Data management
● Data loading/ingestion
● Queries/search
● Use cases
● API/extensibility
● Transactions
● High availability and
disaster recovery
● Deployment options
Full
Potential of
the Property
Graph Model
The Rise of the Graph Database
2010 - 2020: From Model to Category
29. Graph Data Science is a science-driven
approach to gain knowledge from the
relationships and structures in data,
typically to power predictions.
Rise
of(Graph)
Data
Science
What is Graph Data Science?
30. Rise
of(Graph)
Data
Science
o Get more insight out of the
data you already have.
o Add relationships into your
existing workflows.
o Make better, more accurate
predictions.
Why Do Graph Data Science?
31. Neo4j Graph Data
Science Library
Scalable Graph
Algorithms & Analytics
Workspace
Native Graph
Creation & Persistence
Neo4j
Database
Visual Graph
Exploration
& Prototyping
Neo4j
Bloom
Practical Integrated Intuitive
Rise
of(Graph)
Data
Science
Neo4j for Graph Data Science
33. • Degree Centrality
• Closeness Centrality
• Harmonic Centrality
• Betweenness Centrality & Approx.
• PageRank
• Personalized PageRank
• ArticleRank
• Eigenvector Centrality
• Shortest Path
• Single-Source Shortest Path
• All Pairs Shortest Path
• A* Shortest Path
• Yen’s K Shortest Path
• Minimum Weight Spanning Tree
• K-Spanning Tree (MST)
• Random Walk
• Breadth & Depth First Search
• Triangle Count
• Local Clustering Coefficient
• Connected Components (Union Find)
• Strongly Connected Components
• Label Propagation
• Louvain Modularity
• K-1 Coloring
• Modularity Optimization
• Euclidean Distance
• Cosine Similarity
• Node Similarity (Jaccard)
• Overlap Similarity
• Pearson Similarity
• Approximate KNN
• Adamic Adar
• Common Neighbors
• Preferential Attachment
• Resource Allocations
• Same Community
• Total Neighbors
... Auxiliary Functions:
• Random
graph generation
• Graph export
• One hot encoding
• Distributions & metrics
• Node2Vec
• Random Projections
• GraphSAGE
Rise
of(Graph)
Data
Science
50+ Graph Algorithms in Neo4j
34. Write Your Own
Algorithms
Machine Learning
Algorithms
Use matrix math and neural networks
to learn the structure of your graph
Train, store, and fit predictive
models within Neo4j
Leverage your embeddings to build
nearest neighbors graphs
Use our Pregel API to build your own
algorithms using our infrastructure
Graph Embeddings & Complex
Data Types
Model
Catalog
GDS 1.0
First enterprise
library for graph
data science.
Neo4j 4.0
Compatibility
GDS 1.2
GDS 1.1
Graph Mutability
& Expressivity
GDS 1.3
New Algorithms &
RBAC
Graph Native
Learning
GDS 1.4
Rise
of(Graph)
Data
Science
Introducing GDS 1.4
35. Graph algorithms to uncover
trends and patterns
Patterns
Pointers
Queries to answer questions
with connected data
Predictions
Graph-native ML learns the
topology of your graph to
uncover new facts
Rise
of(Graph)
Data
Science
Putting it all Together
36. NODES talks:
- Graph native learning: introducing GraphSAGE & the model catalog
- Write your own algorithms with the Pregel API
- Getting graph questions answered through Neo4j Bloom
Developers:
- Developer Guides: neo4j.com/developer/graph-data-science/
- GDS Sandbox: sandbox.neo4j.com/?usecase=graph-data-science
- GitHub: github.com/neo4j/graph-data-science
Books:
- Graph Algorithms: Practical Examples in Apache Spark & Neo4j:
neo4j.com/graph-algorithms-book/
- GDS For Dummies: neo4j.com/graph-data-science-for-dummies
Rise
of(Graph)
Data
Science
Neo4j Graph Data Science - Resources