10  dissemination

abstract: dissemination of results. Publication of software and research data for reuse.

The final step in the work process is the dissemination of results, which includes traditional publication methods (papers, books, conference presentations) as well as newer approaches, such as the publication of software used in the process, the generated data, and data and software studies focusing specifically on these, blogging and microblogging, sharing presentation slides and recordings, and participating in professional organizations. I would like to emphasize the importance of special data sharing. Imagine the following scenario: a researcher has worked hard to enrich a popular data source with high research potential that is maintained by a public collection. Later, another researcher would like to use the same database for their research. If she is not familiar with the previous researcher’s work, she can start the data enrichment process from scratch. But even if the first researcher published his data enrichment, it is much more likely that the subsequent researchers will find and use the original database. To prevent this, researchers would need to return the modified data to the original data provider. Fortunately, MARC21, introduced in the 34th update in 20221 a data provenance subfield to distinguish between data recorded by the library and data recorded by the researcher (and it is available in most fields), which could be a theoretical remedy for the library’s legitimate demand to take responsibility for its own data. In the life sciences, researchers can use nanopublications to share data enrichment steps with libraries, which can then incorporate them into their catalogs without compromising their own responsibility and credibility. The second researcher can then work on the data-enriched version. In order to realize this vision, communication between the parties must be standardized, and the research community can play a coordinating role in this process.

The final step in the work process is the dissemination of results, which includes traditional publication methods (papers, books, conference presentations) as well as newer approaches, such as the publication of software used in the process, the generated data, and data and software studies focusing specifically on these, blogging and microblogging, sharing presentation slides and recordings, and participating in professional organizations.

Let’s start with the data. There are some dedicated research data repositories that aim to help researchers to publish their data. Along with the data files you should add metadata such as title, authors, subject headings, description. From repository to repository it might be different what metadata schema you should follow and what are the mandatory and optional metadata. The repositories assign a persistent identifier to your dataset, such as DOI, Handle, or ark.2 We can mention Zenodo, Harvard Dataverse, figshare, Open Science Framework, Dryad as the largest general repositories, but there are a number of others with regional or domain specific focus (at time of writing we are not aware of any that provide extra functionalities for bibliographic data). You can check re3data, the registry of research data repositories, that provides a rich categorisation to find the one that fits your needs. It is also worth it to check if your institution has any researh data management recommendation or policy.

Publishing the software might be a two step process. The first step is to make it publicly available in a platform such as GitHub, GitLab, or other general or institutional software depository with a proper license. However you can go further and turn the scripts into real research software with proper documentation, tests, packaging, installation scripts etc.3 Research data repositories also accept research related software, and some of them are working together with software repositories, so you can connect them together, and you will get a persistent identifier for your software as well. The Research Software Engineering community published some guidelines on how to publish software in FAIR way (FAIR is an acronym for Findable, Accessible, Interoperable and Reusable)4.

Finally, we would like to call attention to the importance of a special data sharing called ‘data roundtrip’.5 Imagine the following scenario: a researcher has worked hard to enrich a popular data source with high research potential that is maintained by a public collection. Later, another researcher would like to use the same database for their research. If she is not familiar with the previous researcher’s work, she can start the data enrichment process from scratch. But even if the first researcher published his data enrichment, it is much more likely that the subsequent researchers will find and use the original database. To prevent this, researchers would need to return the modified data to the original data provider. Fortunately, in the 34th update in 20226 MARC21 introduced a data provenance subfield to distinguish between data recorded by the library and data recorded by the researcher (and it is available in most fields), which could be a theoretical remedy for the library’s legitimate demand to take responsibility for its own data. In the life sciences, researchers can utilize a special, “atomic” data publication method called nanopublications to share data enrichment steps with each other. In our case these ‘others’ are libraries, which can then incorporate them into their catalogs without compromising their own responsibility and credibility. The second researcher can then work on the data-enriched version. In order to realize this vision, communication between the parties must be standardized, and the research community can play a coordinating role in this process.

  1. https://www.loc.gov/marc/up34bibliographic/bdapndxg.html↩︎

  2. For a full list with advantages and disadvantages of the persistent identifiers see Koster, Lukas. 2020. Persistent identifiers for heritage objects. Code4Lib Journal 47. https://journal.code4lib.org/articles/14978↩︎

  3. If you are interested in this process we recommend the following book: Nelson, Catherine. 2024. Software Engineering for Data Scientists: From Notebooks to Scalable Systems. O’Reilly. ISBN 978-1-0981-3620-8 https://www.oreilly.com/library/view/software-engineering-for/9781098136192/↩︎

  4. Beyer, Florian, Vedder, Lucia, Singson, Lea Sophie, Sahwan, Wahib, & Schmidt, Marcus. 2025. Publishing research code FAIR - a roadmap (1.0). Zenodo. DOI 10.5281/zenodo.14772749↩︎

  5. Sandra Fauconnier. 2019. Data Roundtripping: a new frontier for GLAM-Wiki collaborations⁠. https://diff.wikimedia.org/2019/12/13/data-roundtripping-a-new-frontier-for-glam-wiki-collaborations/; Wikimedia Deutschland e.V. 2019. Wikidata Use in Cultural Institutions. Research Report. https://upload.wikimedia.org/wikipedia/commons/e/e1/Research_Report_%E2%80%93_Use_of_Wikidata_in_GLAM_institutions_%282019-11%29.pdf↩︎

  6. MARC 21 format for bibliographic data. Update No. 34, July 2022. https://www.loc.gov/marc/up34bibliographic/bdapndxg.html↩︎