Avoid Predatory Journals

In the Article “Predatory Journals: No Definition, no defence” (Springer Nature, Vol 576, Dec 2019), the Authors Argue that Predatory Journals are a Global Threat. I Agree with that. Without Performing Checks for Scientific Quality and Issues such as Plagiarism, it is easy to Scale the Number of Accepted Articles or even Special Issues; the Resulting Business Model is Based on Establishing open Access broad-discipline Journals, with Article Processing Charges paid by the Authors. In These Journals, Editorial and Academic Rigor is Neglected in Favor of Operational Speed and Business Interests. Interestingly, Publication fees of Predatory Journals are Ultimately paid out by the Funding Agencies such as DFG and BMBF (in Germany).

The Purpose of this text is to Motivate the German AI Community to Ensure that our German Journal of Artificial Intelligence from the GI can keep up its Rigorous Reviewing Process and will Never be Subject to such a Development. We will Hence cross-check our Journal’s Characteristics. Scientific Publishers (Including Springer Nature) have Helped to Establish Guidelines for Editorial Board Members and Authors.

One of these guidelines says that negligent peer review is a prominent feature of predatory journals. Here, the call is on you. You are asked to do something, to help ensure that we get enough quality reviews for our submission. Delayed processing times are most often a result of the fact that it takes time to get at least three meaningful and expressive reviews per submitted article because many colleagues reject to review articles on a regular basis. The acceptance rate, based on the scientific review process, is currently at about 30%. Luckily, Springer Nature supports us in this process with desk reject information such as automatic plagiarism check results and concurrent submission information to ease the task. However, we need your voluntary service, dear experts of AI in Germany, in the same way as the readers and the AI community need voluntary editorial board members for scientific, non-profit publishing. Springer Nature and the GI give us something back which is very precious: if you are a corresponding author affiliated with a German university or research institution, you are entitled to publish open access in our KI Journal with fees covered by the German DEAL agreement.

Another predatory journal guideline highlights the lack of transparency. This concerns transparency in operational procedures (e.g., how editorial decisions are made, publication fees applied, and the organisation of peer reviews), covered by the previous paragraph. A second point concerns contact information or details about article processing charges (cf. DEAL). Predatory journals and publishers often make guest editors and members of their editorial boards unverifiable. In contrast, you can visit our webpage (https://​www.​springer.​com/​13218) to see the KI Journal’s information in a transparent way. A third point about transparency is important to mention: Springer announced “Online First” in 2014, all final articles are published online after an author has reviewed proofs and all corrections have been carried out. They are in citable form 2–3 weeks after acceptance and before distribution of the print journal as a special issue. Metadata is sent to all relevant bibliographic services for inclusion in indexing databases immediately after online publication; here DBLP is very relevant for the AI community. Over the last 10 years, we have established a stable publication schedule of 12–14 accepted articles per issue (please note the indicator for predatory journals, that it is easy to scale the number of accepted articles, as mentioned above, to be in stark contrast to this).

Downloads per geography in 2012 include North America (9%), Asia Pacific (31%), and Europe (56%) compared to downloads per geography in 2023 of North America (16%), Asia Pacific (28%), and Europe (47%). Successful full-text downloads are a very promising development: 8797 (2011), 9894 (2012), 16,887 (2013), …, 92,019 (2020), 114,628 (2021), and 134,858 in 2022. The h5 Index was 13 in 2014 and is 23 in 2023. Most papers have been submitted and accepted by Germany, USA and Denmark, followed by Austria, France, Japan and the UK, making up 85%.

I hope I have motivated to publish with us. The authors in “Predatory journals: no definition, no defence” argue that it can be tough to distinguish a predatory journal from a journal that is under-resourced. I would like to repeat my main point. We need your voluntary service, dear experts of AI in Germany, in the same way as the readers need voluntary editorial board members for scientific, non-profit publishing, which requires collaboration and support. The service of the editorial board members, the guest editors, the reviewers and German AI community and members has already achieved the following: the Scimago Journal Rank (SJR), which is based on Elsevier’s Scopus database, is generated by an independent agency, calculating the number of citations in one year to a journal’s articles in the preceding three years; the KI Journal is in the second quartile (heading towards top quartile). Other academic publication incentives include an impact score of 1,91 and a CiteScore of about 4, matching AI Magazine (referring to the magazine character of our Journal with AI market, interviews, discussions, news, dissertation abstracts and project sections). A cite score of about 4 is also achieved by AI and Society or the Journal of Robotics and Control, emphasising the technical articles’ strength. A recent development is that the KI Journal has been indexed by Clarivate’s Web of Science and the journal’s impact factor (JIF) is being calculated.

Best regards,

Daniel Sonntag