Supporting researchers in the publication process with open, algorithm-based tools

There is a whole range of web-based services out there designed to assist researchers in the publication process - in their role as authors, editors, reviewers - at its various stages. Few of those, however, are inherently aligned with the needs and interests of the research community: the underlying procedures and data sources are not transparent and the source code not publicly available, their sustainability perspective is unclear, they are often provided by commercial actors. In our talk, we present two approaches with which we want to do things differently: B!SON, an open journal recommender that is already available, and our ideas for a reviewer recommender, based on the preliminary work and procedures from the B!SON development.

B!SON is a tool for identifying suitable Open Access journals for a manuscript, developed and pro- vided by TIB (Technische Informationsbibliothek) and SLUB (Saxonian State and University Library Dresden). It uses community-based, open data sources, discloses its recommendation algorithm based on semantic and bibliometric similarity detection, handles user data sensitively and has been designed with sustainability in mind. The simple user interface, with input fields for a manuscript's title, ab- stract and references, provides users with a list of Open Access journals in which similar research has been published. For each journal listed, a similarity score is provided, along with the underlying articles where similarity was found, to make the recommendation verifiable. With the development of B!SON, we primarily aim at supporting researchers in their decision-making process and promote open access publishing options. At the same time, however, we want to support libraries as well as other institutions providing publishing advice, but also journals and publishers, especially smaller ones. With our new project we want to build on the semantic and bibliometric similarity detection of B!SON while applying the same principles of openness in the development, but investigate a different use case: the need for an open tool for reviewer search, especially in the emerging landscape of Open Access journals run by the scientific community. Providing a tool to better manage the peer review process that can be integrated into existing publication infrastructures will support these infrastructures and the publications appearing there, and eventually enhance their reputation and attractiveness. Its use carries the potential to promote a faster, appropriate and less biased review process. In addition to reducing the workload for the selection of reviewers by editors, this can increase the confidence of authors in community-based open access publications and ultimately the quality of these publications. Furthermore, such a tool, which draws on large data corpora of already published literature and other scientific output, makes it easier to find reviewers for innovative and unique as well as inter- or transdisciplinary topics for the respective sub-disciplines than is possible with manual searches, and it can also provide initial indications of possible conflicts of interest.

Presentation for a talk held at MaRDI Minisymposium "Towards a digital infrastructure for mathematical research" on September 25, 2023.