Comparisons

ANDS strongly recommends a central integrated infrastructure or hub, so that institutions can track their data collections as intellectual assets. ANDS has funded the development of various metadata stores software solutions (VIVO, ReDBox, TARDIS). Different metadata store solutions have different ranges of functionality and fill different niches.

ANDS‐funded solutions have not been mandated. Nevertheless, to ensure sustainability, we recommend that the various solutions available are being fit‐for‐purpose and able to be integrated into enterprise systems. To this end, institutions are strongly encouraged to enter into consortial arrangements in order to maximise the value of the available funding.

See this table of comparison [Google doc] to see which 4 universities have chosen to deploy VIVO and which 15 universities have chosen to deploy ReDBox (UQ and the CSIRO are building their own and ANU is using the ORCA software supporting Research Data Australia - RDA).

In the following short video Duncan Dickinson describes what ReDBox does:

Comparison of schematics

• Griffith University - Griffith Research Hub
• University of Queensland - UQ Data Collections Registry
• James Cook University - Research Data Catalogue
• La Trobe University – La Trobe University Metadata Store
• RMIT - Research Data Storage and Curation at RMIT
• University of Melbourne - Research Data Registry
• Deakin University - Research Metadata Store
• University of Adelaide - Research Metadata Store Project
  
• Macquarie University - Institutional Metadata Stores
• University of Sydney - University of Sydney Metadata Store
• University of NSW - UNSW Metadata Stores Project
• University of Tasmania - Tasmanian Research eData Directory Service
• University of South Australia – Research Metadata Store

Griffith University - Griffith Research Hub

Summary: With the Metadata Stores Project funded by ANDS, Griffith eResearch Services will further implement and develop the Griffith Research Hub to build on functionality already implemented in previous ANDS and Griffith funded Projects. Full scale creation, management and workflow of metadata records which will be fully operationalised. eResearch staff will be employed fulltime for 12 months from February 2012 to develop the system. A Project Manager will be employed to work with eResearch Staff, Scholarly information Staff and Office for Research Staff to develop and implement functionality. Some of these staff will be devoting time to the project in-kind. All mandatory deliverables will be met, including an operational feed to Research Data Australia, inclusion of metadata records describing research activities and datasets from numerous Griffith faculties, implementation of Parties and Activities services, an operational workflow, and implementation of robust storage and management of metadata. A number of optional deliverables will also be met as outlined in the proposal.

University of Queensland - UQ Data Collections Registry

Summary: This project will develop, test, configure and deploy a UQ metadata store for research data collections that aggregates metadata from existing sources of truth and feeds metadata to ANDS Research Data Australia, the ARDC Party Infrastructure, and the UQ Operational Data Store. The project will also develop a specification of portal requirements for a UQ Academic Portal, and analyse potential software solutions meeting those requirements.

James Cook University - Research Data Catalogue

Summary: In 2011, the ANDS-funded Tropical Data Hub (TDH) Project (EIF046) saw the establishment of the Tropical Data Hub Aggregator. The TDH provides a central point for enabling researchers to deposit data and metadata, as well as providing RIF-CS records to Research Data Australia. It was always intended that the deliverables of EIF046 would form a starting point for the TDH with further development and integration taking place with subsequent projects. One of the deliverables of the EIF046 was an investigation of Research Data Catalogue candidates and how they would integrate with existing TDH architecture. With the criteria used, it was determined that ReDBox would be the most suitable candidate for a Research Data Catalogue within the TDH, however, how well this software could be integrated with existing JCU policies and procedures was not fully investigated. The initial phase of this project will review this recommendation in relation to the current environment of research data management, as well as the internal operating environment within JCU. Following the selection of a Research Data Catalogue solution, an implementation phase will commence so as to extend the current functionality of the TDH by:

• Providing a broader level of metadata descriptions of research data, including the description of the full range of RIF-CS objects (Collections, Party Records, Activities, Services)
• Integration with JCU’s institutional data resources, the NLA’s ARDC Party Infrastructure Project and the various ANDS data identification resources (e.g. ARCS activities)
• The development of a researcher profile system
• Supplying APIs for machine-to-machine submission of metadata from research tools

JCU will share information and project documentation with the research data management community, as well as share discussions about the development and delivery of the optional deliverables to increase overall coverage and project efficiency.

La Trobe University – La Trobe University Metadata Store

Summary: This project will contribute to La Trobe University’s research data infrastructure by establishing mechanisms to:

• manage descriptions of data collections held at La Trobe
• enable discovery and reuse of data collections held at La Trobe
• support strategic planning for research at La Trobe
• ensure high quality of metadata, in support of the other requirements for a metadata store

Initial investigations suggest that a deployment of the RedBox application will best complement La Trobe University’s “sources of truth” applications which include Research Master and a VITAL/Fedora repository.

RMIT - Research Data Storage and Curation at RMIT

Summary: This project will build on the previous work funded by ANDS and other bodies that has produced a number of domain-specific metadata stores with feeds to the Australian Research Data Commons (ARDC). This ANDS-funded work will be integrated with a major University project to build a comprehensive research data storage and curation system. The focus of this ANDS-funded component will be to: create a central metadata store, provide for populating this with metadata from the implemented domain-specific metadata stores, provide a framework for the inclusion of new domain specific metadata stores, and to provide a mechanism to publish metadata to the ARDC. The broader ITS project will provide an overall architecture, business processes and on-going support, basic services and the deployment of reliable, sustainable and extensible storage infrastructure.

University of Melbourne - Research Data Registry

Summary: The aim of the overall program is to leverage existing work and experience to develop a ‘turn key’ Research Data Registry solution for Australian research institutes that both satisfies institutional requirements for registering the location of research data, but also maximizes the value of this registry by:

• Providing a systematic systems based method to audit compliance with research data management requirements
• Profiling research data sets via a research data portal
• Acting as a harvest point for local institution-based research data repositories
• Acting as a harvest point for national data registries (such as the ANDS Data Commons)
• Providing a staging area for other national and international discipline-based repositories

Phase 1 of the program aims to establish the infrastructure required to create an Institutional Research Data Registry, by creating a reusable solution for developing an institutional harvest point for the ANDS data commons on the VITRO platform.

Deakin University - Research Metadata Store

Summary: Broadly, Deakin aims to deliver an enterprise wide metadata store that meets the requirements of the University, can be embedded sustainably within the University’s processes and meets the requirements set out by ANDS. We aim to develop and entrench processes supporting the ongoing capture and ultimate sharing and reuse of metadata associated with research activities. Firstly we will complete a business requirements study to identify the strategic objectives and business requirements for implementation of an enterprise wide solution (incorporating both infrastructure and processes) for research metadata capture, management and publication. We will then complete a feasibility study to identify, review and assess the main solution options and recommend an appropriate solution. The study will include:

• Review of ReDBox pilot Projects (DC20A, DC20B)
• Identification and assessment of alternate solutions for metadata management , (e.g., Vivo)
• Assessment of key advantages, disadvantages, costs, risks and issues
• Consideration of future support requirements
• Opportunities for integration with existing or developing systems

University of Adelaide - Research Metadata Store Project

Summary: The Research Metadata Store project began in April 2012. Innodev were engaged for the solution delivery. The project was progressed using a phased delivery approach with a traditional waterfall software lifecycle. Governance was provided through a Steering Committee, meeting monthly throughout the project. The project team conducted a comparative review of candidate metadata store solutions; ReDBox – Mint was selected as the software platform. Engagement with the ReDBox community and the ReDBox project team at QCIF occurred throughout the project and the project team acknowledge Duncan Dickinson and Andrew Brazzatti from QCIF for their significant contributions to DataConnect and their support for the project team. Design work was undertaken to accommodate the researcher self-registration workflow in ReDBox via an online form and to integrate the Mint with other University systems. The Mint is populated with data about people and projects from the University’s human resources system (PeopleSoft) and research management system (ResearchMaster). Researchers can use lookup functionality available through the online form to cite people and projects from those sources of truth. Additionally the system references external sources of truth including the Australian Research Data Commons (ARDC) Activity and Party Infrastructures, i.e., the persistent URLs used by Australian Research Council and National Health and Medical Research Council grants (Activities) and the National Library of Australia’s Trove ‘People and Organisations’ identifiers (Parties). Working with these additional system components added layers of complexity to the project giving rise to issues with different applications of various metadata standards across systems; appropriate metadata mapping, and achieving metadata quality whilst maintaining best practice for linked data. Nonetheless the integration achieved with internal sources of truth ensure validity and integrity of information sourced locally, while referencing external sources of truth represents good linked metadata practices. The online form is available to all University researchers for self-registration of their datasets and data collections, however the workflow is specifically triggered for Category 1 grants when they are marked closed in the University’s research management system, at which point an email is sent to the grant’s Chief Investigator alerting them to DataConnect. A video about Data Connect is avaiable here: https://vimeo.com/66046536

Macquarie University - Institutional Metadata Stores

Summary: The Macquarie University Metadata Stores initiative will aggregate the institutional metadata information for research collections. As an aggregator the system aims to interconnect data collection, which have native storage and provide the over-arching meta-information to facilitate discovery and usage of research information. This initiative extends another ANDS project; Seeding the Commons SC11, which is driving management of data and research collections with the pilot implementation of a Metadata store. The pilot has evaluated a number of metadata store software solutions enabling easier decisions around the technology choices. Additionally the pilot has setup many of the policies and procedures for meta data collection within research collections. The project will formalise the implementation, configuration and creation of inter-system connectors for the gathering of data and publication to ANDS.

University of Sydney - University of Sydney Metadata Store

Summary: Leveraging existing ANDS projects in data capture (DC), seeding the commons (StC) and metadata aggregation (MA), as well as existing university administrative systems, the project will provide infrastructure in support of research. It will provide an enterprise-wide research data management solution by bringing together existing components, filling gaps, providing a metadata store for research data collections and a system for managing metadata created through StC. In support of strategic planning, the system will enable reporting on published collections. Metadata management for unpublished collections will not be within scope, but will be recommended for attention post-project.

University of NSW - UNSW Metadata Stores Project

Summary: The UNSW Metadata Store project will deliver an enterprise metadata storage system. Components of the project will include the:

• Specification development for software
• Software development at enterprise level
• Installation, configuration and testing of the software

University of Tasmania - Tasmanian Research eData Directory Service

Schematic of UTas Architecture



UTas Network Topology

Summary: The project proposes to create a Tasmanian Research eData Directory Service which describes key elements of a diverse range of data stores held by Tasmanian research entities. This directory service will interface to national metadata stores and hence provide a more extensive and coherent discovery service to data assets held within the state. Management of the service will be coordinated via the UTAS library, through curation, archival, harvesting, and publishing frameworks. Tasmanian research entities have a long legacy of producing extensive high quality data across a broad spectrum of research disciplines including Physical Sciences – agriculture, fisheries, and ore deposits, marine sciences – Antarctic and south oceans; medical sciences – health populations; Arts. These datasets will form the basis of the curation, harvesting, and publishing frameworks through the diligent identification of key data attributes and the population of these attributes into the metadata store. The TReDDS will also conform and where possible exceed national standards for metadata store interoperability and information exchange – which the long term strategy of contributing toward a national metadata store fabric.

University of South Australia – Research Metadata Store

Summary:
The University of South Australia’s Research Metadata Store Project will implement an enterprise metadata store solution. The project will seek requirements, assess existing sector solutions, specify a software solution, develop or implement a system, and interface to both internal and external research discovery services such as Research Data Australia (RDA). The system will be developed as an enterprise class metadata store that will be an enabler of complete, high quality, compliant and accessible metadata for UniSA’s research activities. The system will facilitate ongoing contribution to various Australian Research Data Commons (ARDC) repositories allowing discovery, reporting and reuse of UniSA’s research activities. The project will be directed by UniSA’s central IT unit in close collaboration with Library and Research units. UniSA will also seek collaboration from Adelaide and Flinders Universities locally, and the wider sector, to obtain key learning opportunities for code sharing where possible.