application modelling with graph databases
DESCRIPTION
Applications are built around domain and business concepts. As developers, we model these concepts and their relationships all the time in our heads, on whiteboards, and in code. Then we perform mental gymnastics translating objects and relationships into tables, rows and columns. Wouldn't it be better if our data storage thought the same way we do? This talk will describe a few types of problems that relational databases and many contemporary NOSQL solutions have trouble modeling, and how graph databases are a solution to those problems. Along the way, attendees will be introduced to the graph database Neo4j, will see how to interact with it via the Cypher query language, and learn how they can start modeling their own application domains as graphs.TRANSCRIPT
Application Modelingwith Graph Databases
@josh_adellhttp://www.servicetrade.com
http://blog.everymansoftware.com
http://github.com/jadell/neo4jphp
https://joind.in/10430
The Problem
The Solution?> -- First degree> SELECT actor_name FROM cast WHERE movie_title IN (SELECT DISTINCT movie_title FROM cast WHERE actor_name='Kevin Bacon')
> -- Second degree> SELECT actor_name FROM cast WHERE movie_title IN (SELECT DISTINCT movie_title FROM cast WHERE actor_name IN (SELECT actor_name FROM cast WHERE movie_title IN (SELECT DISTINCT movie_title FROM cast WHERE actor_name='Kevin Bacon')))
> -- Third degree> SELECT actor_name FROM cast WHERE movie_title IN(SELECT DISTINCT movie_title FROM cast WHERE actor_name IN (SELECT actor_name FROM cast WHERE movie_title IN (SELECT DISTINCT movie_title FROM cast WHERE actor_name IN (SELECT actor_name FROM cast WHERE movie_title IN (SELECT DISTINCT movie_title FROM cast WHERE actor_name='Kevin Bacon'))))
The TruthRelational databases aren't very good with relationships
Data
RDBMs
Try again?
Warning: Computer Science AheadA graph is an ordered pair G = (V, E)
where V is a set of vertices and
E is a set of edges,which are pairs of vertices in V.
If vertex pairs in E are ordered,the graph is directed.
Property Graph
Nodes have properties and labels
Relationships have properties, a type and direction
Relationships are first-class entitiesQueried just like Nodes
Indexes
Unique constraintsnew in Neo4j 2.0!
Graphs are Everywhere
Relational Databases are Graphs!
Everything is connected
Modeling "Whiteboard-Friendly"
Nouns => nodes, Verbs => relationships
Back to Bacon
MATCH p = shortestPath( (r:Actor) - [*] - (b:Actor) )
WHERE r.name=”Keanu Reeves” AND b.name=”Kevin Bacon”
RETURN p, LENGTH(p)/2
SocialMATCH
(:Person {name:"Josh"})-[:FRIEND_OF]-(p:Person),(m:Movie)
WHERE NOT(p)-[:HAS_WATCHED]->(m)
RETURN COUNT(p) as not_seen, m
ORDER BY not_seen DESC LIMIT 1
But Wait...There's More!Mutating (insert, update ~ create, merge)
Indexing (auto, full-text, spatial)
Batches and Transactions
Embedded (for JVM) or REST
Where fore art thou, RDB?Aggregation
Ordered data
Truly tabular data
Few or clearly defined relationships
Questions?
Resources● http://github.com/jadell/neo4jphp
● http://neo4j.org
● Jim Webber - A Little Graph Theory for the Busy Developer○ http://vimeo.com/76713692 - Jim Webber
● http://joshadell.com
● https://joind.in/10430