"Finding data is serendipity and luck"
Not surprisingly, the habits and needs varied widely depending on researchers' disciplines and research processes. But what all had in common was that they were struggling with access, findability and quality of data.
How can I find relevant information? Many researchers find it hard to get a good overview of relevant data, because access to information is widely restricted by a paywall, data protection or format. Another issue many face is choosing the right keyword combinations to find relevant information in a search. Many felt, that there are a lot of datasets in Switzerland, and abroad, that would be useful to have access and linkage to but that scientific information tends to be decentralised and difficult to find.
Which related topics of research activities exist in my research domains? One researcher went as far as claiming "finding data is serendipity and luck" due to missing digital infrastructure but also an unwillingness to share data within the research communities. Consequentially, they also claimed, to often be unaware of their research peers working on related topics, even when peers are producing data that could benefit their research. Another researcher expressed: "It would be great to know, who has used datasets, what other datasets where they linking them to and so on."
Is this dataset trustworthy? When asked about re-using data, many researchers mentioned a lack of trust in the quality of data they find. "Quality of data is hard to define and obtain”, they stated, "and is not even solved for domain-specific quality metrics. Best-practices sometimes exist, but I believe greater care with metadata and the methodology used to generate data would be vital."
Ways to find, explore and organise research data in the Connectome
The Connectome aims to provide the user with an either exact, related or approximating set of relevant scientific information that match a given need during the research process. Finding scientific information currently requires search functionalities that point towards a larger set of potentially relevant information, that users need to narrow down through various filtering methods. However, “finding” does not necessarily need to be based on search functionalities alone. Varying strategies can also lead to discovery: such as constructing or pre-constructed queries, monitoring the latest dataset changes and selecting sources.
Among the core Discovery features that will be designed and evaluated during the pilot phase are "Search and Filter" and the more exploratory "Query Generator" (e.g. Query My Social Networks), "Recommendation" or "Monitor" features. In contrast to the traditional "Search and Filter" functionalities, the "Query Generator" is a designed workflow to guide a user through the process of identifying relevant scientific information from the most relevant sources based on their profiles and relationships. Another way to discover potentially relevant scientific information is the personal "Recommendation" for a user.
A second aspect that will be designed and validated with the researchers in the prototype are features to organise data: to create, save, edit, share, import and export collections of scientific information obtained.
Next steps
Based on the results of the interviews and evaluated use cases a partner agency will design and prototype user interfaces. Once these prototypes are developed, the interviewed researchers will test the use cases and the user journeys. The aim is to have a solid base with the pilot partners to define a Minimum Viable Service for the Connectome by November 2020.
