Graph and RDF databases

Graph and RDF Databases
Context : Course of Advanced Databases
Prepared by : Nassim BAHRI
February 19th, 2015

Table of contents
I. Introduction :Overview of BIG DATA & NOSQL
II. Graph Databases
III. RDF Databases
IV. Application example
V. Scientific article
VI. Conclusion and Q&A

Introduction
3
Not Only SQL
Storing data in
memory
Distributed
databases

Introduction : Data Model
4
Documents Databases
(Voldemort, Riak)
Big Table Column
(Hbase, cassandra, Hypertable)
Key-Value
(MongoDB)
Graph Databases
(Neo4J)

Introduction : Data Model
5
Data complexity
Datasize
Key-Value Stores
Column Family
Document Databases
Graph Databases
90% of use cases
This is what we
are interested
Source : Neo Technology webinar

Graph Databases
What is Graph Database?
A graph database is a databases whose
specific purpose is the storage of graph-oriented data
structures.
 Is simply an object oriented database based on Graph
theory.
6

Graph Databases
Representation
• Nodes
• Relationships between nodes
• Properties on both
7
2
3
1
Name : John
Age : 43
Name : Google
Type : Ford
Color : blue
Work in
Since : 2013

Graph Databases
The power of Graph Databases
Performance Flexibility
Agility
8

Graph VS Relational Databases
Relational Database Modeling
ID Name
1 Larry Page
2 Sergey Brin
3 Larry Elisson
N …
ID Name
1 Google
2 Oracle
… …
N …
PersonID CompanyID Since
1 1 1998
2 1 2001
3 2 2010
Person
Company
WorksIn
SELECT Person.Name
FROM Person,Company,WorksIn
WHERE Company.Name='Google'
AND WorksIn.CompanyID=Company.ID
AND WorksIn.PersonId=Person.ID;
Google's employees?
Lookup
Lookup
Lookup
9

Graph VS Relational Databases
Graph Database Modeling
Name : Larry Page Name : Google
Name : Sergey Brin
Name : Oracle
Name : Larry Elisson
Person 1
Person 2
Company 1
Company 2
Person 3
WorksIN
Since : 2001
Since : 2010
Since : 1998
Lookup
10

Graph Databases
Graph storage and graph processing
1. The underlying storage
• Some databases use native graph storage,
• The other databases use relational database, an object-oriented database,…
2. The processing engine
• The nodes are physically connected to each other in database,
• index-free adjacency
11

Graph Databases
Graph Database Management System
12Source [1]

Graph Databases : Example
Visual Modeling
13
Name : John
Age : 27 FRIEND_OF
Name : Sally
Age : 32
Title : Graph Databases
Authors : Ian Robinson,
Jim Webber
Since : 01/09/2013
Since : 01/09/2013
On : 02/09/2013
Rating : 4
On : 02/03/2013
Rating : 5
FRIEND_OF

Create a simple dataset
// Create Sally
CREATE (sally:Person { name: 'Sally', age: 32 })
// Create John
CREATE (john:Person { name: 'John', age: 27 })
// Create Graph Databases book
CREATE (gdb:Book { title: 'Graph Databases',
authors: ['Ian Robinson', 'Jim Webber'] })
// Connect Sally and John as friends
CREATE (sally)-[:FRIEND_OF { since: 1357718400 }]->(john)
// Connect Sally to Graph Databases book
CREATE (sally)-[:HAS_READ { rating: 4, on: 1360396800 }]->(gdb)
// Connect John to Graph Databases book
CREATE (john)-[:HAS_READ { rating: 5, on: 1359878400 }]->(gdb)
14

Simple selection from node:
Query 1 : How old are Sally?
MATCH (sally:Person { name: 'Sally' })
RETURN sally.age as sally_age
16

Simple selection from node:
Query 2 : Who are the authors of Graph Databases?
MATCH (gdb:Book { title: 'Graph Databases' })
RETURN gdb.authors as authors
17

Selection using relationship:
Query 3 : Who are sally's friends?
MATCH (sally)-[r:FRIEND_OF]-(person)
RETURN person.name as sally_friend
18

Selection using relationship and group function:
Query 4 : What is the average rating of Graph Databases?
MATCH (gdb:Book { title: 'Graph Databases' })
MATCH (gdb)<-[r:HAS_READ]-()
RETURN avg(r.rating) as average_rating
19

Using order and limit in query:
Query 5 : Who Read Graph Databases First, Sally or John?
MATCH (people:Person)
WHERE people.name = 'John' OR people.name = 'Sally'
MATCH (people)-[r:HAS_READ]->(gdb:Book { title: 'Graph
Databases' })
RETURN people.name as first_reader
ORDER BY r.on
LIMIT 1
20

Visual Modeling
21
Name : John
Age : 27 FRIEND_OF
Name : Sally
Age : 32
Name : Alain
Age : 19
Since : 01/09/2013
Since : 01/09/2013
FRIEND_OF
Since : 01/11/2014

Completing our schema
// Create Alain
CREATE (alain:Person { name: 'Alain', age: 19 })
// Connect Sally and Alain as friends
MATCH (alain:Person { name: 'Alain' })
CREATE (sally)-[:FRIEND_OF { since: 1358818400 }]->(alain)
22
Alain
Sally
John
GDB book

Node / relationship navigation:
Query 6 : Which is shared between Alain and John Friend?
MATCH (alain:Person { name: 'Alain' })
MATCH (john:Person { name: 'John' })
MATCH (alain)-[:FRIEND_OF]-(person)-[:FRIEND_OF]-(john)
RETURN person.name as friend
23

Update node’s properties:
Query 7 : Change Alain name to Larry
MATCH (n { name: 'Alain' })
SET n.name = 'Larry'
Query 8 : Remove property
MATCH (n { name: 'Larry' })
SET n.name = NULL
Query 9 : Add property
MATCH (n { name: 'John' })
SET n += { hungry: TRUE , position: 'Entrepreneur' } 24

RDF Databases
The principle of the web
25
HTTP Request
HTTP Response
URL : http://website.com
Communication protocol : HTTP
Representation language : HTML

RDF Databases
Changing status
26
URL URI IRI
Uniform Resource
Locator
Uniform Resource
Identifier
International Resource
Identifier
http://website.com http://animals.com#lion http://‫.الحيوانات‬tn#lion

RDF Databases
W3C Standards
27
Identification
Representation
Query Reasoning

RDF Databases
RDF
Resource
Description
Framework
28
means
: pages, person, animals
Idea,…
: attributes, characteristics,
Relationship,…
: Model, language and
syntax to build description

RDF Databases
Data model & syntax
Description : (Subject, Predicate, object)
“example : doc.html is created by John and belongs to the music
theme”
29
Doc.html is created by John
Doc.html belongs to music theme

RDF Databases
Data model & syntax
(Subject, Predicate, object)
(Vertex, edge, Vertex)
30
John
Doc.html
Music
Author
Theme

RDF Databases
Labeled graph with URI and literals
31
http://www.website.com/john#me
http://www.website.com/doc.html
Music
http://www.website.com/schema#author
http://www.website.com/schema#theme

RDF Databases
RDF Syntaxes
XML, Turtle, TriG, JSON-LD,…
Turtle syntax
@prefix rdf : <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
@prefix site : <http://www.website.com/schema#>
<http://www.website.com/doc.html>
site:author <http://www.website.com/john#me>;
site:theme "Music".
32

RDF Databases
SPARQL Protocol And RDF Query Language
• Syntax similar to SQL
SELECT data,
FROM data source
WHERE { conditions }
33

RDF Databases
?x rdf:type ex:Person
Get all person
SELECT ?subject ?property ?value
WHERE { ?subject ?property ?value }
Get the full Graph database
SELECT ?x WHERE
{ ?x rdf:type ex:Person .
?x :name ?name . }
Get all person who have a name
34

RDF Databases
Declaring prefixes
PREFIX esen : <http://esen.tn#>
SELECT ?student
WHERE {
?student esen:registeredAt ?x.
}
35

RDF Databases
Optional pattern
PREFIX foaf : <http://xmlns.com/foaf/0.1>
SELECT ?person ?name
WHERE {
?person foaf:homepage <http://john.info> .
OPTIONAL { ?person foaf:name ?name .}
}
name : unbound
36

RDF Databases
Union
PREFIX foaf : <http://xmlns.com/foaf/0.1>
SELECT ?name
WHERE {
?person foaf:name ?name .
{
{?person foaf:homepage <http://john.info> .} UNION
{?person foaf:homepage <http://paul.info> .}
}
} 37

RDF Databases
Minus
PREFIX ex : <http://website.com#>
SELECT ?person
WHERE {
{ ?person rdf:type ?type }
MINUS { ?person rdf:type ex:student }
}
38

RDF Databases
Use case : rich snippets Google
39
<div xmlns:v="http://rdf.data-vocabulary.org/#"
typeof="v:Person">
My name is 
Pierre Dumoulin.
My personal homepage:
<a href="http://www.example.com" rel="v:url" >
www.homepage.com</a>I’m living is

12 street name
city name
,XY
12345.

</div>

Application example (RDF)
Data storage
# Default graph (stored at http://example.org/foaf/aliceFoaf)
@prefix foaf: <http://xmlns.com/foaf/0.1/>
. _:a foaf:name "Alice" .
. _:b foaf:mbox <mailto:bob@work.example> .
. _:a foaf:mbox <mailto:alice@work.example> .
Query
PREFIX foaf: <http://xmlns.com/foaf/0.1/>
SELECT ?name
FROM <http://example.org/foaf/aliceFoaf>
WHERE { ?x foaf:name ?name }
40
Name
Alice
Result

Application example (Neo4J)
Question : Who is older, Sally or John?
41
Name : John
Age : 27 FRIEND_OF
Name : Sally
Age : 32
Name : Alain
Age : 19
Since : 01/09/2013
Since : 01/09/2013
FRIEND_OF
Since : 01/11/2014

Application example (Neo4J)
Who is older, Sally or John?
MATCH (people:Person)
WHERE people.name = 'John' OR people.name = 'Sally'
RETURN people.name as oldest
ORDER BY people.age DESC
LIMIT 1
42

Scientific article
Title : Querying RDF Data from a Graph Database Perspective
Book title : The Semantic Web: Research and Applications
Pages : 346-360
Online ISBN : 978-3-540-31547-6
Series Volume : 3532
Publisher : Springer Berlin Heidelberg
Copyright : 2005
Authors : Renzo Angles
Claudio Gutierrez
43

Scientific article
MODEL LEVEL DATA
COMPLEXITY
CONNECTIVITY TYPE OF DATA
Network physical simple high homogeneous
Relational logical simple low homogeneous
Semantic user simple/medium high homogeneous
Object-O logical/physical complex Medium heterogeneous
XML logical medium medium heterogeneous
RDF logical medium high heterogeneous
44
Table 1 : Summary of comparison among different database models

Scientific article
PROPERTY G G+ GraphLog Gram GraphDB Lorel F-G
Adjacent nodes +/- √ √ √ +/- √ +/-
Adjacent edges +/- √ √ √ +/- √ +/-
Degree of a node X √ √ x ? X x
Path √ √ √ √ √ √ √
Fixed-length Path √ √ √ √ √ √ √
Distance between two nodes X √ √ X ? x x
Diameter x √ √ X ? x X
45
Table 2 : Support of some graph database query languages for the example graph properties

Scientific article
PROPERTY RQL SeRQL RDQL Triple N3 Versa RxPath
Adjacent nodes +/- +/- +/- +/- +/- +/- X
Adjacent edges +/- +/- +/- +/- X x X
Degree of a node +/- x x x x x X
Path x x x x X x +/-
Fixed-length Path +/- +/- +/- +/- +/- X +/-
Distance between two nodes x x x x x x X
Diameter x x x x x x x
46
Table 3 : Support of some current RDF query languages for some example graph properties

Conclusion
• Using Graph database for storing data in graph form or in
hierarchical tree structure.
• Graph database : Performance, Agility, Flexibility
• The shortest path
47

Bibliography
[1] Ian Robinson, Jim Webber, and Emil Eifrem.«Graph Databases».O’REILLY, 2013.
[2] Serge Miranda, Fabien Gandon. «Des Bases de Données à Big Data». Course at
Nice Sophia university, MOOC, 2015.
[3] Michel Domenjoud. «Bases de données graphes : un tour d’horizon». Available
on <http://blog.octo.com/bases-de-donnees-graphes-un-tour-dhorizon> (consulted
18/02/2015).
[4] Neo4J community. «Cypher Query Language». Available on
<http://neo4j.com/developer/data-modeling/> (consulted 18/02/2015).
[5] Frank Manola, Eric Miller, Brian McBride. «RDF 1.1 Primer». Available on
<http://www.w3.org/TR/2014/NOTE-rdf11-primer-20140225/> (consulted
18/02/2015).
[6] Eric Prud'hommeaux, Andy Seaborne. «SPARQL Query Language for RDF».
Available on <http://www.w3.org/TR/rdf-sparql-query/> (consulted 18/02/2015).
[7] Neo4J community. «Introduction to graph databases webinar». Available on
<http://www.neo4j.org/learn/videos_webinar> (consulted 18/02/2015).
48

Graph and RDF databases

Recommended

More Related Content

What's hot (20)

Viewers also liked (12)

Similar to Graph and RDF databases (20)

Recently uploaded (20)

Graph and RDF databases

Editor's Notes