OA-RJ Open Access Repository Junction

Project Methodology

To realise the discovery role the project will design a junction tool to identify deposit targets. The junction will use a number of data sources and mechanisms will deduce which repositories are associated with an item. This information will be made accessible in a format that can easily be adopted by other service providers and interested parties

To address the delivery role we will develop a standalone broker which will enable content providers to deposit materials with multiple recipients. Without a broker, each depositor would need to maintain credentials and a working agreement with each possible target – a system that simply cannot scale. Using a broker aids depositors and recipients because the various parties only need to establish their own credentials with the broker and trust the broker to handle all the technical issues for communication.

The OA-RJ broker design is based on the following model (shown above):

Discovery via the Junction

1. Data object or API query is received at the Broker URI
2. For API queries, the broker is passed user information (IP address, hostname or organisation name) and will return repository details
3. For data objects, the broker will initially receive a package, unpack it, then determines where it belongs following certain deduction criteria.

Delivery via the Broker

1. Depositors (single clients or organisations) pass content to the broker.
2. The broker queries the junction to find target repositories.
3. The data objects are repackaged and sent to the target repositories in a format they can accept via the SWORD protocol.