Australian National Data Service (ANDS) Registry Software scope is expanding rapidly following with a higher focus on relationships and external connections on different records. This causes a long-term problem with ANDS existing software infrastructure, and increase in the volume of input metadata to ANDS registry leads to a time consuming and computationally expensive process. Metadata retention is currently one of ANDS Registry Software biggest problem because the metadata changes over time can create huge difficulty in keeping the enriched metadata up to date.
ANDS Registry aims to leverage existing technology developed by the Research Data Switchboard to increase the capability of handling relationship types and solve our existing problems with metadata retention. Also to provide ANDS partners with relationship visualisation that come bundled with the Research Data Switchboard software stack. As an added bonus, the improved technology stack will provide ANDS with better capabilities for retrieving different form of relationship models, this could bring a sizable improvement to the Suggested Dataset functionality as well as any relationship related projects down the track.
We plan to tackle this issue in a 3-stage process
1. Deploy a local instance of graph database leveraging the RD-Switchboard source code and import our relationship data. This will solve our existing problems with retrieving relationship metadata and improve our import workflow
2. Synchronise the local graph database with an existing Research Data Switchboard database that holds additional relationships and entities not present in our database. This adds invaluable external connections, adds more context to existing data and provides a broader interconnectivity between different ANDS future projects.
3. Provides a visualisation widget that can be integrated to Research Data Australia. This adds to the functionality of Research Data Australia with a new way to discover relevant metadata records that are not available before.
For further information about this project, please contact Amir Aryani .
- 3422 reads