Research Data Management Meets Software Engineering: NFDIxCS Workshop at SE25
Software engineering research and research software engineering are areas of interest due to the vast quantity and importance of research software in both domains. However, several open questions remain among the archival, retrieval, and replication of research results and software. To discuss these issues, the NDIxCS working workshop RDMxSE brought together people from research and industry at SE25 at KIT Karlsruhe.
In the first part of the one-day workshop program on the opening day of SE25 on Febraury 24, various presentations were given on the Open Research Knowledge Graph (ORKG) and the Research Data Management Container (RDMC). All presentations and speakers are available on the event website. Individual contributions such as the presentation Creating a Reusable Executive Environment for WiKoDa have been and will be published on the NFDIxCS blog.
The afternoon was dedicated to discussing the presented topics and further aspects. As a first focus, the participants explored the special features of Artifact Evaluation (AE), in which software artifacts are evaluated beyond established paper reviews. Questions relating to the quality evaluation of software, useful criteria for badges, their archiving and establishment were discussed in detail. Furthermore, the special features of software in their executability compared to text and data publications were emphasized and it was discussed whether the software or the paper should have priority in the publication process.
The second area of focus revolved around the representation of software and data in knowledge graphs. The differentiation here is challenging; all the more so when code is increasingly generated on the basis of data with the help of AI tools. As the RDMC combines metadata for both software and data, new requirements arise for their description along the question of what is genuinely specific to data and what is specific to software. While it was made clear that the major difference lies in the executability of software, it also became apparent that there is still no semantic concept for differentiating between data and software in a combined container environment and that there is a need for further research in this area.
The annual Software Engineering (SE) conference serves as a platform for exchanging experiences and insights in the field of software engineering. The conference is initiated by the Software Engineering division of the German Informatics Society (Gesellschaft für Informatik e.V.). Numerous members contribute to the success of the event with their expert contributions. The conference is aimed at both software developers from industry and researchers from academia.
