A guest blogpost by Nick Wise
http://www.eng.cam.ac.uk/profiles/nhw24
The Hindawi journal Wireless Communications and Mobile Computing is booming. Until a few years ago they published 100-200 papers a year, however they published 269 papers in 2019, 368 in 2020 and 1,212 in 2021. So far in 2022 they have published 2,429. This growth has been achieved primarily by the creation of special issues, which makes sense. It would be nearly impossible for a journal to increase its publication rate by an order of magnitude in 2 years without outsourcing the massive increase in workload to guest editors.
Recent special issues include ‘Machine Learning Enabled Signal Processing Techniques for Large Scale 5G and 5G Networks’ (182 articles), ‘Explorations in Pattern Recognition and Computer Vision for Industry 4.0’ (244) and ‘Fusion of Big Data Analytics, Machine Learning and Optimization Algorithms for Internet of Things’ (204). Each of these special issues contains as many papers as the journal published in a year until recently. They also contain many papers that are flagged on Pubpeer for irrelevant citations, tortured phrases and surprising choices of corresponding email addresses.
However, I am going to focus on one special issue that is still open for submissions, and so far contains a modest 62 papers: ‘AI-Driven Wireless Energy Harvesting in Massive IoT for 5G and Beyond’, edited by Hamurabi Gamboa Rosales, Danijela Milosevic and Dijana Capeska Bogatinoska. Given the title of the special issue, it is perhaps surprising that only two of the articles contain ‘wireless’ in the title and none contain ‘energy’. The authors of the other papers (or whoever submitted them) appear to have realised that as long as they included the buzzwords ‘AI’, ‘IoT’ (Internet of Things) or ‘5G’ in the title, the paper could be about anything at all. Hence, the special issue contains titles such as:
- Analysis Model of the Guiding Role of National Sportsmanship on the Consumer Market of Table Tennis and Related IoT Applications
- Evaluation Method of the Metacognitive Ability of Chinese Reading Teaching for Junior Middle School Students Based on Dijkstra Algorithm and IoT Applications
- The Construction of Shared Wisdom Teaching Practice through IoT Based on the Perspective of Industry-Education Integration
Of the 62 papers, 60 give Hamurabi Gamboa Rosales as the academic editor and 2 give Danijela Milosevic. Why is the distribution of labour so lopsided? One can imagine an arrangement where the lead editor does the admin of waving through irrelevant papers and the other 2 guest editors get to say that they’ve guest-edited a special issue on their CV.
Of course, in addition to boosting publication numbers for the authors and providing CV points for the guest editors, every paper in the special issue has a references section. Each reference gives someone a citation, another academic brownie point on which careers can be built. An anonymous Pubpeer sleuth has trawled through the references section of every paper in this special issue and found that Malik Bader Alazzam of Amman Arab University in Jordan has been cited 139 times across the 62 papers. The chance that the authors of almost every article would independently decide to cite the same person seems small.
The most intriguing fact about the papers in the special issue however, is that only 4 authors give corresponding email addresses that match their affiliation. These 4 include the only 3 papers with non-Chinese authors. Of the other 58, 1 uses an email address from Guangzhou University, 6 use email addresses from Changzhou University, and 51 use email addresses from Ma’anshan University. All of the Ma’anshan addresses are of the form 1940XXXX@masu.edu.cn and many are nearly sequential, suggesting that someone somewhere purchased a block of sequential email addresses (you do not need to be at Ma’anshan University to have an @masu email address). The screenshot below shows a sample (the full dataset is linked here).
A subset of the titles from the special issue with their corresponding email addresses, all of the form 1940XXXX@masu.edu.cn |
The use and form of the email addresses suggests that all of these papers are the work of a paper mill. It is hard to imagine otherwise how 51 different authors could submit papers to the same special issue using the same institutional email domain and format. Indeed, before 2022 only 2 papers had ever used @masu.edu.cn as a corresponding address according to Dimensions. It is equally hard to imagine how Hamurabi Gamboa Rosales is unaware. How can you not notice that, of the 19 papers you receive for your special issue on the 12th of July, 18 use the same email domain that doesn’t match their affiliation? This may also explain why Hamurabi has dealt with almost all the papers himself. This special issue should be closed for submissions and an investigation begun.
Stepping back from this special issue, this is not an isolated problem. There are at least 40 other papers published in Wireless Communications and Mobile Computing with corresponding emails from Ma’anshan, and Dimensions finds there are 46 in Computational Intelligence and Neuroscience, 38 in Computational and Mathematical Methods in Medicine and 30 in Mobile Information Systems, all published in 2022 and all in Hindawi journals. What are the chances that 18404032@masu.edu.cn is used in a special issue in Computational Intelligence and Neuroscience, 18404038@masu.edu.cn in Disease Markers and 18404041@masu.edu.cn in Wireless Communications and Mobile Computing?
Finally, masu.edu.cn is only one example of a commonly used email domain that doesn’t match the author’s affiliation. It is conceivable that the entire growth in publications of Wireless Communications and Mobile Computing, Computational Intelligence and Neuroscience (163 articles in 2020, 3,079 in 2022) and Computational and Mathematical Methods in Medicine (225 in 2020, 1,488 in 2022) is from paper mills publishing in corrupted special issues.
Nick Wise
*All numbers accurate as of the 12th October 2022.
Yes, I stupidly agreed to review a paper on sports nutrition and ant colonies submitted to the special issue of ‘Explorations in Pattern Recognition and Computer Vision for Industry 4.0’ In my naivety I thought it might be an investigation of human behaviour regarding sports nutrition choices compared to interactions seen in ant colonies. It turned out to be the gibberish produced by an infinite group of monkeys on typewriters. Any reasonable editor should have spotted that in a minute. I will never get those hours back.
ReplyDeleteI always thought Hindawi was a standard “predatory” journal and was amazed to see it affiliated somehow with Wiley. What’s gone on there? Has the Hindawi “leopard changed its spots”, or has Wiley dropped its guard in the pursuit of Open Access easy pickings?
ReplyDeleteThe latter. Springer and Elsevier have also travelled similar routes. So have IEEE, ACM, IoP, etc.
DeleteThe Infinite Monkey Optimisation Algorithm doesn't get enough attention.
ReplyDeleteAh, the Special Issue business. I like to imagine Oprah "you get a SI to edit!" "and you get a SI!"
ReplyDeleteWhat I see from my naive academic colleagues is that they started to buy their own BS: guest editing for MDPI/Hindawi/Frontiers is their humble contribution to the field and to open access. Hint: it is not.
Funny - the last paper in this special issue was accepted on October 12... the day this post was published.
ReplyDeleteIf you observe the current published article most of them are special issues .
ReplyDeleteOn the other hand most normally submitted current papers review process have been taking more six month. Is this journal plans to stop publishing regular paper. If the review process take more than 6 months, why people submit their article to this journal (2500$ APC ).