Discover and use Linked Open Data

The first pilot sprint of the Research Data Connectome has been successfully completed. Architecture, a knowledge graph and a design prototype with four user-validated use cases show how researchers could search for, discover and organise linked data better in the future.

Testo: Sebastian Sigloch, pubblicato il 23.11.2020

The pilot partners created a design prototype for a discovery service based on the Research Data Connectome in response to the needs of researchers and service providers as future users. The goal that unites all of the sprint’s work packages is to jointly establish and model an ontology for the Connectome Knowledge Graph and to construct the Linked Data Pipeline as an interface between data providers (such as DaSCH and FORS) and the Connectome platform.

Development of the design prototypes

Twenty-one researchers in varying disciplines from 11 institutions and three pilot partners were surveyed in semi-structured interviews. The goal of these interviews was to identify common problems and needs in the research data life cycle.
Fictitious users (personas), workflows, and user journeys for a generic discovery service ‘AskReco’ were then developed based on these findings and assessments of use cases. The researchers participated in the entire prototyping process and were consulted during various stages of the development process.

In the image above, for example, a fictitious user, the sociologist Clementine Schmidt, is searching for datasets relating to the subject ‘Chinese Cities Population Database’ and gets the result shown. Along with the known results, Clementine also receives additional relevant information from the knowledge graph such as ‘7 publications are based on this dataset’ or insights such as ‘Reco identified linkages to 77 other works from 3 different disciplines’. These simple examples already demonstrate the added value of Linked Open Data. The displayed relevance score of ‘84’ refers to personally specified preferences for the search algorithm (for example ‘Search Keywords included in Title, Abstract and Descriptions’). The lower image shows different kinds of ‘collections’ (personal collections, group collections, recommended collections).

All use cases and user processes were validated by the researchers: the surveyed researchers tested all the individual user interface prototypes. The results were used to develop improved design wireframes.

Roadmap 2021

The goal in the coming year is to expand the service portfolio and partner network for use of the data from the Research Data Connectome. In the next step, the design wireframes will be transferred to an operational discovery platform. Additional features can be defined, tested and developed depending on the needs of various stakeholders.

Partners will also be acquired for specific services (e.g. discipline- or application-specific, such as data science). Our goal is to define new features for these services together with our stakeholders and develop them together with our partners.

Sull’autore
Sebastian   Sigloch

Sebastian Sigloch

Sebastian Sigloch joined SWITCH in February 2015 as Innovation & Business Development Manager. He concluded a Doctor of Philosophy research on Internet Economics in Cambridge and graduated in Business Development at École Superieure de Commerce in Clermont-Ferrand and Business Information Technology at Reutlingen University.

E-mail

The Research Data Connectome 

Scientists across disciplines generate increasing amounts of valuable data as part of their daily research activities. Being able to reuse or even combine such scientific data opens the door to many exciting possibilities. Until now, research data has been collected in domain or institutional silos and could not be easily connected.

The Research Data Connectome connects and organises (open) scientific (meta)data sustainably across disciplines to make it widely accessible, interoperable and valuable. Building a Connectome prototype is a joint effort by DaSCH, FORS, EPFL Blue Brain, eXascale Infolab, SATW,
SAGW and SWITCH.

Altri contributi