Introduction

Graph database is a structured database, a database management system that differs from the classic model of the relational database management system (RDBMS). Graph database uses nodes, edges, and properties to represent and manipulate data in a schema-less structure. An appropriate methodology should be designed and implemented to setup a database, insert and manipulate data, as well as reporting on the data and data structure. This graph database must support multiple users, in a client-server model and it should allow and support following features,

• To store graph data object into files in uniform format (the format can be BSON, but yet to be decided in Technical Design)
• To provide features of querying, traversing graph data objects stored in programming API mode (these features may be enhanced to a GUI-based application if possible)
• Support data management of insert, update and delete in local-server mode (the remote-connection with fixed protocol may be implemented in application-layer protocol, Restful web service)
• Run the graph database in distributed mode including load-balancing and fail-over clustering (this may be further enhance to bulk synchronous parallel and map reduce)
• Apply index algorithm on object attribute (node property) in graph database (the algorithm can be B* Tree but it is yet to be decided)
• Based on what we have learnt about database and our working experience, we have decided the primary programming language to be used in this project - Java. Furthermore, we will make use of free development tools, such as Eclipse. NetBeans; and existing open source components, such as data indexing library, RDF parser library and BSON parser library to develop our functional graph database.

Please contact web admin if you have any inquiry:

Project Admins:

• Chang, Xiaoran