Lead: Mag. Thomas Haselwanter (University of Innsbruck)
The focus of the work package is on developing a model approach to selecting an institutional repository for research data. The procedure is demonstrated on a selection process for a long tail repository for the University of Innsbruck and general recommendations for such a selection process are given. Specific repositories are not recommended because each system has its own opportunities and risks. Rather, it is the responsibility of the research organization to choose an appropriate system in weighing opportunities and risks.
Based on a total of 50 semi-structured interviews with 57 researchers from the University of Innsbruck and UMIT Hall from all faculties, the need for a repository for research data was identified in 2019. Since the selection of a long-term system is very important for a research institution, various information was used for the process. The technical requirements were checked by means of the criteria catalog of the Research Data Alliance (RDA). In addition, persona and use cases for research data repositories have been developed to better understand the needs and objectives of the stakeholders involved.
The success of an institutional repository can be deduced from its use by the persons involved. Therefore, it has been a particular concern to evaluate ease-of-use through hands-on tests by researchers, library staff and the central IT services.
The work package tested the three open source repositories Dataverse, D-space, Invenio, and four commercial repositories of ExLibris, Elsevier, Figshare, and TIND.IO. First, the open source repositories and commercial repositories were presented to researchers from various disciplines, members of the library, and the Central Informatic Services. This was followed by a two-week trial period of commercial repositories at the University of Innsbruck. The open source repositories and commercial repositories were evaluated against the criteria of the RDA. In addition, the commercial repositories were also evaluated for usability on the basis of a short questionnaire (see Zhang 2013 ).
The procedure for selecting a repository for research data has proven to be workable and expedient and is recommended. Materials created and used in the work package (including personal descriptions, application scenarios) will be publicly available and applicable to the own selection process of other research organizations. Regardless of the choice of repository, guidelines for filing and storing research data are recommended for the researchers. Different types and versions of data (raw data, master data, sensitive data, citable data, etc.) can be stored, shared and / or published in the repository or in a sync & share system.
Deliverables:
Haselwanter, Thomas & Thöricht, Heike (2020). Creation of personas for the selection of an institutional repository for research data. Digital Library of the University Innsbruck. DOI 10.25651/1.2020.0001
Haselwanter, Thomas & Thöricht, Heike (2020). Use cases for research data repositories. Digital Library of the University Innsbruck. DOI 10.25651/1.2020.0002
Haselwanter, Thomas & Thöricht, Heike (2020). The storage process of research data and what can be learned from Zenodo. Digital Library of the University Innsbruck. DOI 10.25651/1.2020.0004
Haselwanter, Thomas & Thöricht, Heike (2020). Classification of research data and storage systems. Digital Library of the University Innsbruck. DOI 10.25651/1.2020.0005