This is the second Part of Maarten van Kampen grand opus which began with a single special issue at the Elsevier journal Engineering Analysis with Boundary Elements (EABE), and then went all over the place because there is so much papermill fraud going on.

The second part is dedicated to one German man: Timon Rabczuk, the Vice-president of the Bauhaus University of Weimar. Again, Maarten was helped by Alexander Magazinov and Tu van Duong.

This was Part I:

The Bauhaus University refused all communication.

---

Karimipour Saga: Vietnamese Bauhaus

By Maarten van Kampen

In Part I of the Karimipour Saga we discussed a series of fraudulent special issues. The last special issue in the series, Recent trends and new developments in Molecular Dynamics and Lattice Boltzmann Methods, was published in the journal Engineering Analysis with Boundary Elements (EABE) and is a cesspit of fraud. The special issue was originally edited by Arash Karimipour and Amir Mosavi, but the name of that last editor suddenly disappeared from the EABE website somewhere in November last year.

The first special issue in the series, Recent Advances in the Modeling of Nanotubes within Nano-Structures / Systems, was edited by amongst others apex-papermiller Ali J. Chamkha and our current subject, Timon Rabczuk.

Rabczuk’s name also popped up in Part I in connection with a Saudi-Arabian affiliation scam. We now make a detour from Karimipour to take a look at the work of Rabczuk, showing him to be very well-connected to the actors of Part I.

Rabczuk is currently running a special issue that is still open for submission. And its papers will be published in… EABE. As is more often the case, EABE is confused about its special editors: Rabczuk’s name is missing as editor on the actual call for papers, bottom-right:

The above special issue can be used to make a few more bridges to part I. One of its papers, An intelligent approach to investiate the effects of container orientation for PCM…, is authored by Hakan Öztop. Öztop earned a small part in this FBS piece on Ali J. Chamkha, Rabczuk’s co-editor on the Act I special issue. He currently has 8 papers flagged on Pubpeer. Two of these earned an Expressions of Concern because they were published in the Act 2 special issue [1, 2]! Finally, a quick glance at his EABE special issue paper showed significant self-plagiarism, see PubPeer.

But back to Rabczuk:

Curriculum Vitae

Timon Rabczuk‘s CV can be found on his webpage at the Bauhaus University of Weimar:

Rabczuk received his engineering education in southern Germany, concluding it with a PhD in 2002 at the Fraunhofer Institute in Freiburg. After this he went twice abroad, first for a 3 year post-doc position at Northwestern University (Illinois, USA), and later for a 2 year stay in New Zealand. In 2009 he settled back in Germany, receiving a full-time professorship at the Bauhaus University of Weimar. In the past 15 years he has made his career there and since June 2023 he can call himself Vice President of Research and Projects of Bauhaus University.

The most interesting part about a CV is the information that is not there. So let’s dig a tiny bit deeper. Rabczuk’s ResearchGate profile shows a nice list of affiliations:

The Bauhaus affiliation we obviously recognize. But what about the rest that are completely missing from his CV?

King Saud

Let’s start with the King Saud University in Saudi Arabia. The Spanish newspaper El Pais revealed last year that Saudi universities were paying up to 70,000 dollar/year to have ‘Clarivate highly cited’ scientist commit affiliation fraud. And Rabczuk happens to be one of the 12 German professors this investigation identified, see our Friday Shorts. It is often not easy to tell the main affiliation of a researcher, with Rabczuk’s affiliation list above above a good example of the complications. Clarivate took to simply asking the listed researchers for their affiliation. And as a result corrupted scientists could get loads of Saudi money by, well, lying about their affiliation.

In the case of Rabczuk his Clarivate record looks like this:

In 2014 Rabczuk was first listed as highly cited researcher and for the first three years he was content with his Bauhaus affiliation. In 2017 and 2018 King Saud became his primary affiliation, with Bauhaus still second. And in the 2019-2021 period he moved fully to Saudi Arabia, dropping the secondary attribution to “his” Bauhaus university. In 2022 Rabczuk no longer made it into Clarivate’s list of Highly Cited Researchers, possibly because of being delisted as he appears cited in equal measure in 2022.

The Saudi bit is completely missing from his CV, which just shows a full-time employment as professor at Bauhaus University. And in his publication record on OpenAlex we can see that only in 2018 he started signing a fraction of his papers with a (virtual) King Saud affiliation: 0% (2017), 11% (2018), 19% (2019), 8% (2020), and 3% (2021).

Affiliation trickery like the above is often very much unappreciated by the university that is paying for its full time professor. Take the case of this Dutch professor Vincenzo Fogliano who was also feeding from the King Saud trough: it took just two weeks before he had to relinquish control of his group.

[Bauhaus University is aware, but their silence indicates they are happy with their Vice President. Maybe his governance colleagues hope for Saudi affiliation for themselves? -LS]

Double Vietnam

In Rabczuk’s affiliation list we also find two Vietnamese universities: Duy Tan and Ton Duc Thang University. And you can probably feel where this is going: like Karimipour, Rabczuk has a keen interest in advancing Vietnamese science from the goodness of his heart.

Due to this charitable behavior he became the subject of a flattering article in a major Vietnamese newspaper, aptly headlined “Lying about affiliations to promote university rankings“. The authors address the stellar rise of Vietnamese universities in the Shanghai Ranking. They show that it is just a bubble inflated by contributions from non-Vietnamese researchers who sign their papers with Vietnamese affiliations. The article singles out a number of authors:

And yes, the author first mentioned is our Rabczuk. And number two is… his wife Xiaoying Zhuang. Zhuang is incidentally also Rabczuk’s most frequent co-author, featuring on more than 25% of his ~940 papers! And she is not only professor in Tongji University in China, but she is also Professor for Mechanics and Chair of Computational Science and Simulation Technology at the University of Hannover in Germany.

Zhuang Xiaoying of Tongji Uni. granted Sofja #Kovalevskaja Award, one of Germany's most valuable research prizes pic.twitter.com/gnXLalIwAT

— China Xinhua News (@XHNews) November 18, 2015

Number 3 in the list is the papermill fraudster Shahaboddin Shamshirband, whom Alexander Magazinov described as “one of the most retracted researchers ever, for over forty of his “research pieces“:

In the ‘Lying about affiliations’ newspaper article it is noted that the authors are affiliation-fluid: they publish with a range of affiliations and sometimes even change affiliation between submission and publication. Take the below Rabczuk paper:

In the SciHub ‘Accepted’ version Rabczuk is affiliated with Duy Tan University and his actual Bauhaus employer. But in the as-published version he became affiliated with… Ton Duc Thang University. Is there maybe a spot market for affiliation fraud, with authors rapidly switching affiliation based on fluctuating rewards?

The above affiliation fraud is in itself already very bad, but it gets even worse. In the acknowledgements of the paper one can find the following:

The authors acknowledge support by the European Union through the the ERC COMBAT project. This was a project run by Bauhaus University that got a 2 million euro grant from the EU between 2014 and 2019. And guess who the principal investigator is:

Yes, Rabczuk is the principal investigator. Our PI professor is happily accepting 2 million euro of European money, whilst at the same time optimizing his return-on-investment on a Vietnamese affiliation scam and cashing in on his ‘highly cited researcher’ accolades by also posing as King Saud researcher. And no, this was not a single mistake. There are 32 papers listed on PubPeer that were funded by Rabczuk’s EU project, but where he was using his more profitable affiliations in Saudi Arabia and Vietnam:

It is also worth noting that between 2014 and 2019, this Saudi-Arabian academic was participating in another €2 million-heavy EU grant: the Horizon 2020 project named “Environmentally best practices and optimisation in hydraulic fracturing for shale gas/oil development (BESTOFRAC)“. And although Saudi Arabia is fittingly oil-rich, he decided to author his first paper on the project’s publication list with a Vietnamese affiliation:

Rabczuk’s Vietnamese affiliation milking is, however, far more professional than that of Arash Karimipour. For one, he received the prestigious Ton Duc Thang University Price for the excellent rankings that the university obtained. And there are photographs of the price giving ceremony that show that Rabczuk has actually visited Vietnam!

I find it interesting that none of his CVs mention this brilliant award, not even the ‘Honors and Awards’ section of this one. What is also rather astonishing is that none of his CVs mention his second full professorship at Ton Duc Thang University:

A full professorship does not seem a small thing. Imagine your CV: Bauhaus full professor, Bauhaus Vice President, and Ton Duc Thang full professor. Let’s hope his Tongji, King Saud, and Duy Tan affiliations are virtual or fake, otherwise Rabczuk would never have time for some well-deserved vacation…

A highly cited paper

Clarivate’s Highly Cited Researchers list is filled with two types of researchers: famous influential bigwigs, and outright frauds. In citation lists, also garbage floats up because fraud is effective, making these lists highly enriched with dubious types. I want to discuss here one of Rabczuk’s highly cited papers. And it actually comes in a trio:

Date	Title	Authors
2020.06	COVID-19 Pandemic Prediction for Hungary; A Hybrid Machine Learning Approach	Gergő Pintér, Imre Felde, Amir Mosavi, Pedram Ghamisi, Richard Gloaguen
2020.10	COVID-19 Outbreak Prediction with Machine Learning	Sina Faizollahzadeh Ardabili, Amir Mosavi, Pedram Ghamisi, Filip Ferdinand, Annamária R. Várkonyi-Kóczy, Uwe Reuter, Timon Rabczuk, Peter M. Atkinson
2020.11	Coronavirus Disease (COVID-19) Global Prediction Using Hybrid Artificial Intelligence Method of ANN Trained with Grey…	Sina Faizollahzadeh Ardabili, Amir Mosavi, Shahab S. Band, Annamária R. Várkonyi-Kóczy

The topmost paper was submitted first and is published in MDPI’s Mathematics. It is one of these bombshell papers that takes the scientific community by surprise: it has accrued more than 220 citations, making it Mathematics‘ most-cited paper of 2020 and its second-most cited paper of all time. And then came Rabczuk’s paper, published in MDPI’s Algorithms. It also caused a splash, becoming Algorithms‘ most-cited paper of 2020 and its fifth-most cited paper of all time. And these accolades did certainly not go unnoticed, with MDPI’s Algorithms proudly posting the achievements on their Twitter/X account:

So did MDPI really struck gold here?

In the author list we see on all three papers a very familiar name: Amir Mosavi! We met him in Part I as the second special editor of Karimipour’s EABE special issue. Apart from being its editor he also turned out to be bulk citation beneficiary of that same SI.

Fittingly, Mosavi used to be sponsored by the German government as “Green Talent” and Alexander von Humboldt Fellow for his activities as Research Fellow at Bauhaus University Weimar and at the Thuringian Institute of Sustainability in the neighbouring town of Jena. Yes, really: our EABE special editor Mosavi was employed at Rabczuk’s own Institute of Structural Mechanics (ISM) at the Bauhaus University.

In the author list of the 2020.11 paper we also see a certain “Shahab S. Band“. That name may not directly ring a bell, but it is the new stage name of Shahaboddin Shamshirband whom we met above in the Lying about affiliations article. Shamshirband drew lots of negative attention on Retraction Watch, and after scoring some 41 retractions he was in dire need of a less recognizable name. Our Rabczuk has published 14 papers with Shamshirband and ~20 with Amir Mosavi. We even have this paper where all three sit cozily together, with Mosavi sporting his Bauhaus affiliation:

Finding Mosavi and Shamshirband as authors on these COVID papers is obviously a bad omen. And what also does not help is that the first two papers only share two authors. Both papers use identical machine learning models that output identical curves. Which raises prickly questions about author contributions.

Since it is mean and ad hominem to discredit papers based on their authors, we will also take a look at their content. And we can fortunately do that in an academic and peer-reviewed way by citing the January 2024 MDPI Algorithms (really!) paper by Károly Héberger. It is titled Frequent Errors in Modeling by Machine Learning: A Prototype Case of Predicting the Timely Evolution of COVID-19 Pandemic, and that mentioned ‘prototype case’ happens to be… Rabczuk’s COVID paper. The author uses strong imagery in his work, namely that of a sick horse:

Take this quote from the introduction (highlights mine):

“One of its figures includes a sick horse, on which all the external and internal horse diseases are indicated that can affect the animal during its life (Figure 1). Naturally, such an animal cannot exist; still, it is used figuratively in the Hungarian language (“állatorvosi ló”, i.e., “veterinary horse”). One cannot find a scientific article that commits all possible (known) errors, but a recent, highly cited paper [2] can be used as a prototype for illustrative purposes.“

Károly Héberger (10.3390/a17010043)

With Ref. [2] being Rabczuk’s highly-cited COVID product. Since this is an academic discourse, the author has to be careful and polite in his wording. Take this snippet (highlights mine):

“The authors honestly admit the fallacious modeling in the discussion section, but then they conclude that “machine learning [is] an effective tool to model the outbreak”. Why? Why did the editor and the reviewers not notice this? Even less understandable is the next statement that occurs twice in the paper: “a genuine novelty in outbreak prediction can be realized by integrating machine learning and SEIR models”, especially if we see that no such integration was involved in the study.

The only plausible explanation for such contradictions can be that neither the editor nor the reviewers read the manuscript fully, if the authors. The authorship is to be determined by the Vancouver criteria as prescribed by the ethical guide of MDPI“

Károly Héberger (10.3390/a17010043)

The author politely makes the point that the paper makes false claims, that neither editor, reviewers, nor authors read the manuscript, and that at least some of these things go against authorship and ethical guidelines of MDPI. On For Better Science we can be far more explicit: the three papers are not only incompetently written and erroneous, they are pure fakes. These stronger concerns were posted on PubPeer some half a year earlier and the editors of both Algorithms and Mathematics were informed in May 2023. When reaching out at the end of January 2024, I learned that their investigation was still ongoing.

All three papers state to use a machine learning algorithm to predict future COVID case numbers based on a limited number of historical data points. The claim is that these machine learning algorithms do this accurately and generalized, without the need to choose a specific extrapolation function. If that is not enough, the neural networks are optimized with nature-inspired “Grey Wolf” and “Imperialist Competitive” algorithms. In Rabzcuk’s paper the general machine learning algorithms are contrasted with the limited effectiveness of fitting predefined functions. Except that the machine learning algorithms of the first two papers exactly reproduce one of the authors’ predefined (logistic) fits for each and every case. And in the last paper the machine learning models magically change their mind and output a pure quadratic function. And that is just lazy fabrication.

There are many fun observations to make about the Rabzcuk’s paper:

• It was posted to 12 preprint servers that often did not match the topic. The one posted to BodoArXiv (History of the middle ages) got even retracted, or rather, deleted completely.
• The authors submitted the paper to eight preprint servers after the paper was published in the open access journal Algorithms. Creating a novelty “postprint” publication?
• Many of the (bought?) citations went to these preprints, with e.g. the SSRN version thus far receiving 64 citations. Without preprints the main paper would have received more than 100 citations extra, making it actually the 3^th most cited Mathematics paper of all time.
• Rabczuk’s paper was the only one in the MDPI special issue “Feature Papers in Evolutionary Algorithms and Machine Learning” (archived link). Amir Mosavi was obviously special editor, together with three other authors on the paper. MDPI must have been properly embarrassed as the above SI disappeared some time after me reporting the paper.
• The authors of the 2020 paper cite a 2013 article in the Bangladeshi International Journal of Agriculture and Crop Sciences. The journal features on Beall’s list and since 2018 its webpages serve porn (not suited for work). [85] Reshadsedghi, A.; Mahmoudi, A., Detection of almond varieties using impact acoustics and artificial neural networks. Int. J. Agric. Crop Sci. 2013, 6, 1008.
• The authors were hell-bent to get the above paper cited. The same paper got also published in Modern applied physics by the publisher Canadian Center of Science and Education. This publisher incidentally also features on Beall’s list, but is still operational. We can therefore still appreciate this important work, although it now comes with different authors: [84] Khalesi, S.; Mahmoudi, A.; Hosainpour, A.; Alipour, A. Detection of Walnut Varieties Using Impact Acoustics and Artificial Neural Networks (ANNs). Mod. Appl. Sci. 2012, 6, 43. And no, the paper was not relevant.
• The author contributions state that Timon Rabczuk contributed to the software, validation, formal analysis, investigation, resources, and data curation. But something must have gone terribly wrong with the software and validation of this highly-cited paper, as the paper’s authors failed to find the best logarithmic fit (PubPeer).

In conclusion, Rabczuk is a highly cited author who combines a Vice President of Research position with two full professorships in different countries. He receives his money from (EU) funding and it is not really credible that he does not receive money for his Saudi-Arabian, Vietnamese, and Chinese affiliation scams. And he is author on an extremely highly cited paper that is utterly fraudulent.

---