Intro to Graphs and Neo4j

(c) Neo Technology, Inc 2014
Graph Database
Introduction
Meetup

April 2014
Michael Hunger
michael@neotechnology.com
@mesirii
@neo4j

Agenda
1. Why Graphs,Why Now?

2. What Is A Graph, Anyway?

3. Graphs In The Real World

4. The Graph Landscape

i) Popular Graph Models

ii) Graph Databases

iii)Graph Compute Engines

Why Graphs?

The World is a Graph

Some Use-Cases

Social
Network

(Network)
Impact
Analysis

Route
Finding

Recommenda<ons

Logis<cs

Access
Control

Fraud
Analysis

Securi<es
&
Debt

What Is A Graph,
Anyway?

A
Graph
Node
Relationship

Four Graph Model
Building Blocks

Property
Graph
Data
Model

Nodes

Rela<onships

Rela<onships
(con<nued)
Nodes
can
have
more

than
one
rela<onship
Self
rela<onships
are
allowed
Nodes
can
be
connected
by
more

than
one
rela<onship

Labels

Four
Building
Blocks
๏ Nodes

• En<<es

๏ Rela<onships

• Connect
en<<es
and
structure
domain

๏ Proper<es

• AJributes
and
metadata

๏ Labels

• Group
nodes
by
role

Whiteboard

Friendlyness
Easy to design and model

direct representation of the model

Tom Hanks Hugo Weaving
Cloud Atlas
The Matrix
Lana
Wachowski
ACTED_IN
ACTED_IN
ACTED_IN
DIRECTED
DIRECTED

name: Tom Hanks
born: 1956
title: Cloud Atlas
released: 2012
title: The Matrix
released: 1999
name: Lana Wachowski
born: 1965
ACTED_IN
roles: Zachry
ACTED_IN
roles: Bill Smoke
DIRECTED
DIRECTED
ACTED_IN
roles: Agent Smith
name: Hugo Weaving
born: 1960
Person
Movie
Movie
Person Director
ActorPerson Actor

Aggregate vs.
Connected Data-Model

What is NOSQL?
It’s not “No to SQL”
It’s not “Never SQL”
It’s “Not Only SQL”
NOSQL no-seek-wool n. Describes ongoing
trend where developers increasingly opt for
non-relational databases to help solve their
problems, in an effort to use the right tool for
the right job.

NOSQL
Relational
Graph
Document
KeyValue
Riak
Column
oriented
Redis
Cassandra
Mongo
Couch
Neo4j
MySQL
Postgres
NOSQL Databases

31

31
Living in a NOSQL World

31
Volume ~= Size

31
Density~=Complexity
Volume ~= Size

31
Density~=Complexity
Volume ~= Size
Key-Value
Store

31
Density~=Complexity
Column
Family
Volume ~= Size
Key-Value
Store

31
Density~=Complexity
Column
Family
Volume ~= Size
Key-Value
Store
Document
Databases

31
RDBMS
Density~=Complexity
Column
Family
Volume ~= Size
Key-Value
Store
Document
Databases

31
RDBMS
Density~=Complexity
Column
Family
Volume ~= Size
Key-Value
Store
Document
Databases
Graph
Databases

31
RDBMS
Density~=Complexity
Column
Family
Volume ~= Size
Key-Value
Store
Document
Databases
Graph
Databases
90%
of
use
cases

31
Aggregate Oriented
RDBMS
Density~=Complexity
Column
Family
Volume ~= Size
Key-Value
Store
Document
Databases
Graph
Databases
90%
of
use
cases

“There is a significant downside - the whole approach works
really well when data access is aligned with the aggregates, but
what if you want to look at the data in a different way? Order
entry naturally stores orders as aggregates, but analyzing
product sales cuts across the aggregate structure. The
advantage of not using an aggregate structure in the database
is that it allows you to slice and dice your data different ways
for different audiences.
!
This is why aggregate-oriented stores talk so much about map-
reduce.”
Martin Fowler
Aggregate Oriented Model

The connected data model is based on fine grained elements
that are richly connected, the emphasis is on extracting many
dimensions and attributes as elements.
Connections are cheap and can be used not only for the
domain-level relationships but also for additional structures
that allow efficient access for different use-cases. The fine
grained model requires a external scope for mutating
operations that ensures Atomicity, Consistency, Isolation and
Durability - ACID also known as Transactions.
!
Michael Hunger
Connected Data Model

Relational vs. Graph
34

You know relational
34

You know relational
34
users

You know relational
34
users skills

You know relational
34
users skillsuser_skill

You know relational
34
now consider relationships...

35

Looks different, ﬁne.Who cares?
35

๏a sample social graph
35

•with ~1,000 persons
35

๏average 50 friends per person
35

๏pathExists(a,b) limited to depth 4
35





๏caches warmed up to eliminate disk I/O
35





35
# persons query time
Relational database 1.000 2000ms





35
Neo4j 1.000 2ms





35
Neo4j 1.000 2ms
Neo4j 1.000.000 2ms

Neo4j is a Graph Database

• A Graph Database:

• a schema-free labeled Property Graph

• perfect for complex, highly connected data

• reliable with real ACID Transactions

• scalable: Billions of Nodes and Relationships, Scale out with
highly available Neo4j-Cluster

• fast with more than 2M traversals / second

• Server with HTTP API, or Embeddable on the JVM

• Server with HTTP API, or Embeddable on the JVM
• Declarative Query Language

Graph Database: Pros & Cons
• Strengths

• Powerful data model, as general as RDBMS

• Whiteboard friendly, agile development

• Fast, for connected data

• Easy to query

• Weaknesses:

• Sharding (they can scale up and out reasonably well)

• Global Queries / Number Crunching

• Binary Data / Blobs

• Requires conceptual shift

• graph-like thinking becomes addictive

Graph Querying

You know how to query a
relational database!

40

Just use SQL
40

Just use SQL
40users skillsuser_skills

Just use SQL
40users skillsuser_skills
select skills.name
from users join user_skills on (...) join skills on (...)
where users.name = “Michael“

How to query a graph?

42

You traverse the graph
42

// find starting nodes
MATCH (me:Person {name:'Andreas'})
Andreas
42

// find starting nodes
MATCH (me:Person {name:'Andreas'})
// then traverse the relationships
MATCH (me:Person {name:'Andreas'})-[:FRIEND]-(friend)
-[:FRIEND]-(friend2)
RETURN friend2
Andreas
42

Cypher
a pattern-matching
query language for graphs

Cypher attributes
#1 Declarative
You tell Cypher what you
want, not how to get it
44

Cypher attributes
#2 Expressive
Optimize syntax for reading
45
MATCH (a:Actor)-[r:ACTS_IN]->(m:Movie)
RETURN a.name, r.role, m.title

Cypher attributes
#3 Pattern Matching
Patterns are easy for your
human brain
46

Cypher attributes
#4 Idempotent
State change should be
expressed idempotently
47

Query Structure

MATCH (n:Label)-[:REL]->(m:Label)
WHERE n.prop < 42
WITH n, count(m) as cnt, 
collect(m.attr) as attrs
WHERE cnt > 12
RETURN n.prop,
extract(a2 in 
filter(a1 in attrs
WHERE a1 =~ "...-.*")
| substr(a2,4,size(a2)-1)]
AS ids
ORDER BY length(ids) DESC
LIMIT 10
Query Structure

MATCH
describes the pattern

WHERE n.prop < 42
WHERE cnt > 12
RETURN n.prop,
extract(a2 in 
filter(a1 in attrs
WHERE a1 =~ "...-.*")
AS ids
SKIP 5 LIMIT 10
MATCH - Pattern

WHERE
ﬁlters the result set

WHERE n.prop < 42
WHERE cnt > 12
RETURN n.prop,
extract(a2 in 
filter(a1 in attrs
WHERE a1 =~ "...-.*")
AS ids
SKIP 5 LIMIT 10
WHERE - ﬁlter

RETURN
returns the result rows

WHERE n.prop < 42
WHERE cnt > 12
RETURN n.prop,
extract(a2 in 
filter(a1 in attrs
WHERE a1 =~ "...-.*")
AS ids
SKIP 5 LIMIT 10
RETURN - project

ORDER BY

LIMIT SKIP
sort and paginate

WHERE n.prop < 42
WHERE cnt > 12
RETURN n.prop,
extract(a2 in 
filter(a1 in attrs
WHERE a1 =~ "...-.*")
AS ids
SKIP 5 LIMIT 10
ORDER BY LIMIT - Paginate

WITH
combines query parts

like a pipe

WHERE n.prop < 42
WHERE cnt > 12
RETURN n.prop,
extract(a2 in 
filter(a1 in attrs
WHERE a1 =~ "...-.*")
AS ids
SKIP 5 LIMIT 10
WITH + WHERE = HAVING

Collections
powerful datastructure
handling

WHERE n.prop < 42
WHERE cnt > 12
RETURN n.prop,
extract(a2 in 
filter(a1 in attrs
WHERE a1 =~ "...-.*")
AS ids
LIMIT 10
Collections

MATCH (:Country {name:"Sweden"})
<-[:REGISTERED_IN]-(c:Company)
<-[:WORKS_AT]-(p:Person:Developer)
WHERE p.age < 42
WITH c, count(p) as cnt, 
collect(p.empId) as emp_ids
WHERE cnt > 12
RETURN c.name AS company_name,
extract(id2 in 
filter(id1 in emp_ids
WHERE id1 =~ "...-.*")
| substr(id2,4,size(id2)-1)]
AS last_emp_id_digits
ORDER BY length(last_emp_id_digits) DESC
SKIP 5 LIMIT 10
Concrete Example

CREATE
creates nodes, relationships
and patterns

CREATE (y:Year {year:2014})
FOREACH (m IN range(1,12) |
CREATE
(:Month {month:m})-[:IN]->(y)
)
CREATE - nodes, rels, structures

MERGE
matches or creates

MERGE (y:Year {year:2014}) 
ON CREATE
SET y.created = timestamp()
FOREACH (m IN range(1,12) |
MERGE
(:Month {month:m})-[:IN]->(y)
)
MERGE - get or create

SET, REMOVE
update attributes and labels

MATCH (year:Year) 
WHERE year.year % 4 = 0 OR
year.year % 100 <> 0 AND
year.year % 400 = 0
SET year:Leap
WITH year
MATCH (year)<-[:IN]-(feb:Month {month:2})
SET feb.days = 29 
CREATE (feb)<-[:IN]-(:Day {day:29})
SET, REMOVE, DELETE

INDEX,
CONSTRAINTS
represent optional schema

CREATE CONSTRAINT ON (y:Year)
ASSERT y.year IS UNIQUE
!
CREATE INDEX ON :Month(month)
INDEX / CONSTRAINT

Graph Query Examples

Social
Recommendation

MATCH (person:Person)-[:IS_FRIEND_OF]->(friend),
(friend)-[:LIKES]->(restaurant),
(restaurant)-[:LOCATED_IN]->(loc:Location),
(restaurant)-[:SERVES]->(type:Cuisine)
!
WHERE person.name = 'Philip' AND loc.location='New York' AND
type.cuisine='Sushi'
!
RETURN restaurant.name
* Cypher query language examplehttp://maxdemarzi.com/?s=facebook

Network Management
Example

Network Management - Create
CREATE !
! (crm {name:"CRM"}),!
! (dbvm {name:"Database VM"}),!
! (www {name:"Public Website"}),!
! (wwwvm {name:"Webserver VM"}),!
! (srv1 {name:"Server 1"}),!
! (san {name:"SAN"}),!
! (srv2 {name:"Server 2"}),!
!
! (crm)-[:DEPENDS_ON]->(dbvm),!
! (dbvm)-[:DEPENDS_ON]->(srv2),!
! (srv2)-[:DEPENDS_ON]->(san),!
! (www)-[:DEPENDS_ON]->(dbvm),!
! (www)-[:DEPENDS_ON]->(wwwvm),!
! (wwwvm)-[:DEPENDS_ON]->(srv1),!
! (srv1)-[:DEPENDS_ON]->(san)!
Practical Cypher

Network Management - Impact Analysis
// Server 1 Outage!
MATCH (n)<-[:DEPENDS_ON*]-(upstream)!
WHERE n.name = "Server 1"!
RETURN upstream!
Practical Cypher
upstream
{name:"Webserver VM"}
{name:"Public Website"}

Network Management - Dependency Analysis
// Public website dependencies!
MATCH (n)-[:DEPENDS_ON*]->(downstream)!
WHERE n.name = "Public Website"!
RETURN downstream!
!
Practical Cypher
downstream
{name:"Database VM"}
{name:"Server 2"}
{name:"SAN"}
{name:"Webserver VM"}
{name:"Server 1"}

Network Management - Statistics
// Most depended on component!
MATCH (n)<-[:DEPENDS_ON*]-(dependent)!
RETURN n, !
count(DISTINCT dependent) !
AS dependents!
ORDER BY dependents DESC!
LIMIT 1
Practical Cypher
n dependents
{name:"SAN"} 6

๏ Full day Neo4j Training & Online Training

๏ Free e-Books

• Graph Databases, Neo4j 2.0 (DE)

๏ neo4j.org

• http://neo4j.org/develop/modeling

๏ docs.neo4j.org

• Data Modeling Examples

๏ http://console.neo4j.org

๏ http://gist.neo4j.org

๏ Get Neo4j

• http://neo4j.org/download

๏ Participate

• http://groups.google.com/group/neo4j

How to get started?
81

ThankYou
Time for Questions!

Intro to Graphs and Neo4j

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Intro to Graphs and Neo4j (20)

More from jexp (20)

Recently uploaded (20)

Intro to Graphs and Neo4j