The Netherlands eScience Center  is the Dutch national center of
excellence for the development and application of research software. We
provide a combination of in-cash and in-kind funding to Dutch research
projects that require a both academic research and software development
to solve their research challenges.
We firmly believe that open software is just as important to open science
as are open access publications and open data. Without open software, the
“instrument” that produces the published results from the data is missing.
In our view, the FAIR principles  should therefore also apply to software; it
needs to be findable, accessible, interoperable, and re-usable. Software
is not the same as data however, as it requires a complex environment to
operate in: compilers, libraries, configuration files, operating systems,
etc. It needs to be maintained to fix bugs and to adapt it to changes
in its environment. In this respect, software is more volatile than data.
As an organization, we want to show that we have a positive impact on
the projects that we do. This is a challenge because our contributions
are mostly in the form of software. For many scientific domains it is
not commonplace to cite software, even when that software is crucial
to the research being presented. Because getting credited is not
typically within our control, we decided to investigate alternative
ways of demonstrating impact of software.
This poster presents the Research Software Directory [3,4],
a content management system that is tailored to software. At the time
of writing, it collects data from GitHub, Zenodo, Zotero (our
organization-wide reference manager), CITATION.cff files (for machine
readable citation data), Medium (our institute's blog), and more.
What little input it requires from our engineers, can be provided through a web form.
Using this information, we create a ‘product page’ for each software
package on the Research Software Directory. Each software package is
presented within its research context, which may include links to
scientific papers, projects, presentations, etc. In addition, it also
shows a social context: who are the developers, how active is the
development, and are there any tutorials, blogs, videos, etc.
Importantly, each product page also clearly states how to cite the
software package in papers.
Besides providing a qualitative assessment of software impact, the
Research Software Directory also improves findability of software
packages, for example by providing metadata that helps search engines
understand what the software is about, but equally importantly, by
providing humans with the necessary context to understand the software's
purpose. After all, discovery of a software package is often not so
much about finding it, but knowing that you found it.