AllegroGraph

From Saltlux

1. Outline
AllegroGraph developed by Franz Corp of U.S.A. is a 64bit RDF Store and application framework for semantic web application construction. AllegroGraph stores triple type data. These data may be queried through various API such as SPARQL and Prolog, and RDFS++ level reasoning is possible through built-in reasoner. For more powerful reasoning full description logic is supported through Racer system. Reasoning is also possible through GUI based tool by connection with ontology builder of TopBraid Composer. AllegroGraph 3.0 version provides new functions of Federation, Social Network Analysis, Geospatial capabilities and Temporal reasoning.

2. AllegroGraph New V3.0 Features
 Federation
 Geospatial Primitives
 Temporal Primitives
 Social Networking Analysis
 Performance improvements

3. Storage and Querying capability


( * Size = size of triple file, **Load Rate = KTPS = Thousand Triples per second )

Hardware - Linux based, dual CPUs (x86-64) 1.8 GHz, with 16 GB RAM. Hardware for Federated is a quad CPU machine with the same specs as dual CPU machine.

On Amazons EC2, AllegroGraph loaded and indexed 10 Billion Triples derived from 1 Billion Telecom CDRs (Call Detail Records) into 10 large instances, 4 parallel loads per instance, in 6.19 hours.

4. Approach method of AllegroGraph Triple-Store
• Java
• Sesame HTTP client protocol
• Python, Ruby, JavaScript, etc
• Lisp
• TopBraid Composer

5. Strong and rich reasoning and querying
• SPARQL Queries on Named Graphs
• RDFS++ Reasoning
• Prolog
• RacerPro and RacerPorter

6. Supplementary functions
• Native Data Types and Efficient Range Queries
• Free-text Indexing
• Named Graphs for Weights, Trust Factors, Provenance
• Direct Reification
• Geospatial and Temporal Reasoning
• Clustering
• Federated Triple Databases

7. AllegroGraph Documentation
• AllegroGraph Documentation
• AllegroGraph FAQ
• AllegroGraph LUBM50 Benchmarks
• AllegroGraph Performance Tuning

8. System requirements


9. Use case
• France Telecom
Having telephone network of 34million users France Telecom developed Recife, a decision making system based on Allegro system and use it.

• British Telecom Lab
Having a world biggest telephone network BTL(British Telecom Lab) used Allegro system to develop SPEED (Pricing and End to End Design) in designing the data network using Allegro system for information storage and management.

• NTT R&D
Japanese telecommunication company NTT R&D has developed ENCORE, an ACL based network processing agent using allegro system for the storage and management system of information. ENCORE enables effective routing between ISP.

• KDDI R&D Lab
Japanese telecommunication company KDDI R&D Lab has developed automatic security management system using Allegro system as for the storage and management system of information.

• Google
Google has used Allegro system as for the storage and management system in may R&D projects.