Tuesday, 22 May 2012

Key Apps Deliverable #1 – RIF-CS descriptions

All software tools developed through the Applications program, the associated input and output datasets used in analyses, and the researchers conducting the analyses on the software tool are expected to be described in Research Data Australia (RDA), ANDS’ discovery service for Australian Research Data.

In order for information to be included in RDA, the information must be encoded in the RIF-CS (Registry Interchange Format - Collections and Services) schema that allows description of these components of Apps projects and the connections between them:
  •          Datasets (which are represented as “Collections” in RIF-CS)
  •          People, Organisations and Institutions (which are represented as “Parties” in RIF-CS)
  •         Software (which are represented as “Services” in RIF-CS)
  •         Funded Research Activities (which are represented as “Activities” in RIF-CS)
The simplest idea to illustrate what is going on in any Apps project can be shown as such…

…where Data (i.e. “Collections”) of different types are being fed into a Software Tool (i.e. a “Service”) to produce new output data (i.e. a new data “Collection”).

Based on this simple information model, the Apps team have produced some example RIF-CS records to describe and illustrate the relationships between various input data sources, the Apps software tool being developed, the output dataset, the researchers using such software products, and the owners of the input and output data. This is illustrated in graphical format here:

The RIF-CS XML for this above example can be downloaded from here

Alternatively, the records and relationships can also be explored using records published to the RDA “Demo” area. The easiest way to explore these sample records in RDA Demo is to start at the service record and to navigate the relationships from one entity to another, backwards and forwards). The service record in RDA can be found here.

Note that the content of these example records are based on pure fantasy, but what we’ve tried to do is to embed a combination of relevant information/instructions/examples at the appropriate places in these example records to demonstrate how to use each RIF-CS element.  Any instructions have been copied directly from the RDA Content Providers Guide.

All of the ANDS Apps projects do need to deliver a set of sample ‘draft’ RIF-CS records in RDA within a couple of months from the project start, that describe the input/output datasets, the software system and the researchers involved in using the system. Your ANDS contact will help you in making these records. Once crafted, the records will be assessed by an ANDS metadata analyst and when approved, the records moved to RDA for public display. The sample records approved at this time are expected to be then used as the model for future RIF-CS records that will be automatically fed to RDA from the finished software product throughout the project and into the future.

No comments:

Post a Comment

ANDS Applications Project Blogs/Sites