README

Setup

• Install Ruby
  • then install Rails: gem install rails
• Install Redis
• Install Neo4j

How to Use?

• Start Neo4j Server with: bundle exec rake neo4j:start
• Start the Application Server with: bundle exec rails s -p 3010

1. Persistence

1.1. Persist triple at Redis and Neo4j at the Same Time:

• Send a POST to http://localhost:3010/insert-data with body:

  { "data": "[[\"John\",\"Is a Friend of\",\"James\"],[\"John\",\"Is a Friend of\",\"Jill\"],[\"James\",\"Is a Friend of\",\"Jesse\"],[\"James\",\"Is a Friend of\",\"Jill\"],[\"Jill\",\"Is a friend of\",\"Doug\"],[\"Jill\",\"Likes\",\"Snowboarding\"],[\"Snowboarding\",\"Is a\",\"Sport\"], [\"Doug\",\"Plays\",\"RPG\"]]" }

  It'll persist this graph in Redis Following the strategy:
  • "subject predicate" => "object" It'll persist this graph in Neo4j Following the strategy:
  • subject:Type -[predicate]- object:Type

1.2. Persist triple at Redis only:

• POST:

http://localhost:3010/save-triple-to-redis

• With BODY:

  { "data": "[[\"John\",\"Is a Friend of\",\"James\"],[\"John\",\"Is a Friend of\",\"Jill\"],[\"James\",\"Is a Friend of\",\"Jesse\"],[\"James\",\"Is a Friend of\",\"Jill\"],[\"Jill\",\"Is a friend of\",\"Doug\"],[\"Jill\",\"Likes\",\"Snowboarding\"],[\"Snowboarding\",\"Is a\",\"Sport\"], [\"Doug\",\"Plays\",\"RPG\"]]" }

It'll persist this graph in Redis Following the strategy: - "subject:predicate" => "object" and predicate => subject:predicate:object

1.3. Persist triple at Neo4j olny:

• POST:

http://localhost:3010/save-triple-to-neo4j

With BODY:

{"data": "[[\"John\",\"Is a Friend of\",\"James\"],[\"James\",\"Is a Friend of\",\"Jesse\"],[\"Jill\",\"Is a friend of\",\"Doug\"],[\"Jill\",\"Likes\",\"Snowboarding\"],[\"Snowboarding\",\"Is a\",\"Sport\"]]" }

It'll persist this graph in Neo4j Following the strategy: - subject:Type -[predicate]- object:Type

The Values stored in Neo4j chan be checked via GUI

• Acess: localhost:7474
  • Run queries
  • MATCH (n:Person) RETURN n LIMIT 25
  • MATCH p=()-[r:IS_A_FRIEND_OF]->() RETURN p LIMIT 25

Observation:

For testing purposes this projects abstracts a graph as an Array like:

 [[subject 0, predicate 0, object 0],
  [subject 1, predicate 1, object 1]]

Tested using:

"[[\"John\",\"Is a Friend of\",\"James\"],[\"John\",\"Is a Friend of\",\"Jill\"],[\"James\",\"Is a Friend of\",\"Jesse\"],[\"James\",\"Is a Friend of\",\"Jill\"],[\"Jill\",\"Is a friend of\",\"Doug\"],[\"Jill\",\"Likes\",\"Snowboarding\"],[\"Snowboarding\",\"Is a\",\"Sport\"], [\"Doug\",\"Plays\",\"RPG\"]]"

2. Querying

• You can test Querying sending a GET request to /parse.
• You should send a SPARQL query as PARAM query.

2.1 Querying Examples:

• GET:

http://localhost:3010/parse?query=SELECT ?p WHERE { ?p IS_A_FRIEND_OF ?x . ?x LIKES 'Snowboarding' }

• For the test data mentioned above, it'll return:

  • John
  • James

• GET:

http://localhost:3010/parse?query=SELECT ?p WHERE { ?p IS_A_FRIEND_OF ?x . ?x PLAYS 'RPG' }

• For the test data mentioned above, it'll return:

  • Jill

• GET:

http://localhost:3010/parse?query=SELECT ?p WHERE { ?p IS_A_FRIEND_OF ?r }

• For the test data mentioned above, it'll return:

  • Jill
  • John
  • James

• GET:

http://localhost:3010/parse?query=SELECT ?p WHERE { ?p IS_A_FRIEND_OF 'Jesse' }

• For the test data mentioned above, it'll return:
  • James