Empirical Analysis of Research Data Search Interfaces and Interaction
The availability and reuse of research data is incresingly becoming a central driver of research in many scientific disciplines. Principles such as the FAIR principles for open research data advocate facilitating findability of existing datasets so that other researchers can reuse them to replicate experiments and extend them. Various domain-specific, institutional or general-purpose research data repositories and registries exist which support registration and search for suitable datasets for a given research problem. Their user interfaces support different search paradigms such as keyword-based, faceted or concept-based approaches. The structure and extent of metadata used to increase findability of existing datasets varies across different platforms and domains. Likewise, the actual search behavior of researchers is not very well understood yet.
The objective of this thesis is to empirically analyze research data search interfaces of established research data platforms and to produce insights on actual use of these interfaces to find suitable datasets. The corresponding research questions focus on the analysis of research data search interface features, search queries and search strategy success. The thesis comprises a literature review of existing platforms and approaches, deriving an abstract model of research data search interface features. For analysis of actual search behavior, empirical experiments with test subjects have to be planned, conducted and analyzed quantitatively and qualitatively. In order to focus on the use of identified features and search behavior and to reduce bias from well-known platforms, layout, visual design etc., suitable uniform representations of the user interfaces of existing platforms need to be created.