You need papers to advance in academia and even in the industry, and at some point it stops mattering what you publish and where these papers originally come from. Enter a gang of electronics engineers from China who decided to expand into multimedia and visual communication research by stealing the work of others they somehow got hold of as peer reviewers and publishing it under their own names. And the names of their friends. Because it worked so well, they published the plagiarised paper again and again, in the same journal. And when the Springer journal had enough and retracted their papers, they simply re-published the same text, slightly garbled, in a different journal with Elsevier.

Now, Springer journal Multimedia Tools and Applications retracted 14 more of these clones in one go, incidentally around same day when Smut Clyde started tweeting about it. Maybe this is why all these 14 retraction notices bear no date? Is Springer that ashamed?

Elsevier on the other hand cannot be bothered, what others perceive as a shameful embarrassment, is a money-making opportunity for the market leader RELX. One needs filler material for journals which nobody seems to read or even to edit, in order to sell subscriptions. And if a Springer paper is retracted, there are no copyright issues standing in the way of republication with Elsevier. Well, maybe there are, because the re-published paper was originally plagiarised, i.e. stolen. In any case, the journal editors seem to have discreetly looked away at all times, and one thing is certain: anti-plagiarism screens are definitely not being applied by these IT experts, like the Journal of Visual Communication and Image Representation Editor in Chief Zicheng Liu from Microsoft.

The following guest post by Smut Clyde and Tiger BB8 illustrates nicely how diseased scholarly publishing has become, to serve perverse career incentives in academia and apparently also in the industry.

---

Springer secretly ashamed, Elsevier lets it all hang out

By Smut Clyde and Tiger BB8

When a journal retracts 14 papers at once – the entire contents of a Special Issue – something is probably happening behind the scenes. Here I may be working the Retraction Watch side of the street, but the story provides enough entertainment for everyone. Also, there are digital doppelgangers and glitches in the Matrix!

The journal in question was Multimedia Tools & Applications, MTAP for short, published by Springer. No doubt this has its devoted readership of experts in multi-format cross-media data integration and analysis, but outside that circle, MTAP is perhaps best known for the ‘peer-review piracy’ episode last year. Authors had submitted manuscripts for editorial consideration, only to see them rejected (after long delays) on the advice of anonymous reviewers, then promptly appear in the same journal as the work of other authors. In fact this happened three times, with the same team of authorial frauds. Quoth Retraction Watch:

Multimedia Tools and Applications (MTAP), has retracted three papers by the same group of authors, all of which plagiarized from unpublished manuscripts by other people.[…]
There are four authors in common to all three manuscripts, but only one — corresponding author Chao Xiong of the Changzhou Institute of Technology in China — has responded to any queries from MTAP, according to the retraction notices.

Now cursory inquiries revealed that Chao Xiong is a specialist in the nano-fields of thin-film deposition and opto-electronics. All credit to him for branching out into a new area, but he seems an unlikely choice for a peer-reviewer, so how he had access to the manuscripts he acquired was a mystery. Any thoughts that there might be two Chao Xiongs at the Changzhou Institute of Technology – one toiling in the groves of nanotech while the other robbed researchers – were dispelled by the presence of Acknowledgements, in the plagiarised papers and in ones by the optoelectronics Chao Xiong, to the same regional research grants. More to the point: the nanotech guy did claim credit for the plagiarisms in his ResearchGate entry.

After Retraction Watch reported the retraction of his three mysteriously plagiarized papers in August 2018, Xiong quieted reduced his internet presence. He removed scrubbed his plagiarism record from his ResearchGate profile, and his latest appearance on internet was from November 2018 when he was mentioned as the Vice Dean of the School of Electrical and Optoelectronic Engineering of Changzhou Institute of Technology. By now also this has changed: Xiong is not listed on the university’s leadership webpage anymore, but still features as an Associate Professor there.

At the time, the MTAP management promised a broader investigation:

“We conducted an extensive investigation and found evidence that the peer-review process for a number of articles in MTAP may have been manipulated. […] As this investigation is still ongoing, we are unable to disclose how the corresponding author of the retracted papers may have obtained unpublished material.”

This recent drainpipe-gurgling swirl of retractions seems to be part of that inquiry, for similar reasons motivated the depublications… combinations of piracy from unpublished manuscripts [presumably submitted to the journal]; false authorship; and falsified peer review. The editorial explanations themselves display a high degree of unacknowledged self-duplication:

RETRACTED ARTICLE: Internet-scale secret sharing algorithm with multimedia applications

“The Editor-in-Chief has retracted this article [1], which was published as part of special issue “Multi-source Weak Data Management using Big Data”, because its content has been duplicated from an unpublished manuscript authored by Her-Chang Chao and Tzuo-Yau Fan without permission. In addition, there is evidence suggesting authorship manipulation and an attempt to subvert the peer review process. Author Chao Xiong agrees to this retraction.”

Retracted on 18.05.2019, according to Retraction Watch database.

RETRACTED ARTICLE: Medical image encryption technique in big media environment

“…because its content has been duplicated from an unpublished manuscript authored by Priya Selvam and Santhi Balachandran without permission. In addition, there is evidence suggesting authorship manipulation and an attempt to subvert the peer review process. Author Chao Xiong agrees to this retraction”

Retracted on 18.05.2019, according to Retraction Watch database.

Now the Chao Xiong team, familiar from last year’s round of thefts, account for only three of the present 14. So what of the other 11? Well, four of them were ostensibly the work of a team at “State Grid Zhejiang Electric Power Company” in Hangzhou, centred on Caiyou Zhang:

RETRACTED ARTICLE: A new deep representation for large-scale scene classification

Bo Dai, Feng Mei. Deliang Ji, Caiyou Zhang, Jia Shi

“…because the article shows substantial overlap most notably with the article cited [2]. In addition, there is evidence of figure duplication without appropriate permission, as well as evidence suggesting authorship manipulation and an attempt to subvert the peer review process.”

Retracted on 18.05.2019, according to Retraction Watch database.

RETRACTED ARTICLE: Cross-camera multi-person tracking by leveraging fast graph mining algorithm

Caiyou Zhang, Yuteng Huang, Zhiqiang Wang, Hongcheng Jiang, Dongfeng Yan, Jun Cheng

“… because part of its content has been duplicated from an unpublished manuscript authored by Her-Chang Chao and Tzuo-Yau Fan without permission. The article also shows substantial text overlap, most notably with the articles cited [2, 3]. In addition, there is evidence of figure duplication without appropriate permission, as well as evidence suggesting authorship manipulation and an attempt to subvert the peer review process.”

Retracted in May 2019, exact date unknown since not listed in Retraction Watch database.

Two more were signed by Zhengwei Jiang, also affiliated to the “State Grid Zhejiang Electric Power Company, Hangzhou”: “Camera network analysis for visual surveillance in industrial electronic context“, and “Analysis of security operation and maintenance system using privacy utility in media environment“, both were retracted on 18.05.2019:

“content has been duplicated without permission from unpublished manuscripts”, “substantial text overlap”, “evidence of figure duplication”, “authorship manipulation and an attempt to subvert the peer review process”.

Another retraction was signed by Yuteng Huang (Huang et al 2017) who has the same employer, while Gui et al 2018 and Zheng et al 2017 were also Zhejiang-based, these were retracted according to Retraction Watch database between 17 and 18 May 2019. As for “Image denoising based on parallel K-singular value decomposition in cloud computing“, retracted on 18.05.2019,

(“there was evidence suggesting an attempt to subvert the peer review process”)

it is reassuring to learn that Hongtao Gao was in the employ of the ‘Department of Cyber Crime Investigation, Criminal Investigation Police University of China, Shenyang’ when he signed it.

Anyway, we need not grieve for these retracted papers, because it is not yet GAME OVER. Even before they disappeared from MTAP with a rude gurgling noise, many had already earned themselves a second incarnation in the Journal of Visual Communication and Image Representation (JVCIR for short), published by Elsevier. Here is a glitch in the matrix; or a case of déjà lu, in Yao et al 2019:

Another glitch, in Jiang 2019:

All four Caiyou Zhang forgeries were reincarnated in JVCIR (sometimes with slight changes in authorship), and both of Zhengwei Jiang’s.

But there’s more! For scrutiny detects papers which have not been taken out from MTAP and shot (yet), but which have already begun their second incarnation in JVCIR.

Oddly, in the doppelganger below, Caiyou Zhang as author of the ‘original‘ Dai et al MTAP 2017 has been replaced with Yiyang Yao in Dai et al JVCIR 2018. The duplicated Abstract has apparently been written by feeding text through a paper shredder and reassembling it, in hommage to the Burroughs-Gysin cut-up technique.

“This article will be visual significance into the graphical guidance (the chart is a medium-sized join subgraph) Deep structure, from the level of learning a significant map The original image pixel to the object level graphic (oGL), and further Space level graphics (sGL). In particular, we first sample Super pixels from each image, and they are used as buildings Block of each object. In order to seamlessly describe different objects The number of oGLs is generated by spatial adjacent links The super pixel oGL object response mapping is obtained by obtaining Transfer the semantics of the image tag to oGL. As space The layout of the object plays an important role in the prominence of the object Based on the relevant learning distribution proposed sGL OGL position between. Finally, in order to imitate the” winner of all” Biological vision mechanism, the largest majority of voting programs The sGL of the image is probabilistically combined into a significant graph. Experimental results show that oGLs and sGLs capture the object level well And space-level visual cues, resulting in competitiveness Significant detection accuracy.”

This process of parallel publication reached its acme with a four-fold paper. This appeared in MTAP – in a different Special Issue, on “Data Security in Multimedia Modeling” in June 2017 – and was retracted on 18.05.2019 for the now-familiar reason of behind-the-scenes piracy:

RETRACTED ARTICLE: Moving object surveillance using object proposals and background prior prediction

Yiyang Yao, Jun Li, Tengfei Wu & Luming Zhang

“…because its content has been duplicated from an unpublished manuscript submitted by a different group of authors. In addition, there is evidence suggesting authorship manipulation and an attempt to subvert the peer review process.”

The main author promptly republished it in February 2018 in the same journal with different co-authors and a different title, and then retracted it in May 2019 (retraction is not recorded in Retraction Watch database):

Efficient object analysis by leveraging deeply-trained object proposals prediction model

Bo Dai, Yiyang Yao, Wenjing Ye & Jing Zheng

Not satisfied that this plagiarised paper had reached its full potential, he went on to recycle both versions  through JCVIR, as and where they appear in a single issue as Dai et al 2019 (Volume 61, May 2019, Pages 218-224) and Dai et al 2019 (Volume 61, May 2019, Pages 85-92).

This is an egregious achievement, even by the standards of academic publishing, and deserves some sort of celebration.

Am I the only one struck by the emphasis in this genre on Big Data mass surveillance, and algorithmic segmentation of video scenes into individuals and backgrounds? It is almost as if a network of academics asked themselves “What sort of Cargo-Cult nonsense should we promise to get research funding from a dystopian government of authoritarians and kleptocrats?”.

Things begin to make more sense – in a sufficiently broad sense of “sense” – when we check back to that MTAP Special Issue on “Multi-source Weak Data Management using Big Data” (a rubric now devoid of contents after all contributions were deemed to be plagiarised, and only accepted because of fake-author shenanigans and bogus peer review).

The name of the supervising Guest Editor should be familiar:

That special issue of MTAP had five guest editors. Among them, the “Leading” guest editor was Yiyang Yao from the State Grid Zhejiang Electric Power Company (SGZEPC) in China, and Luming Zhang from Hefei University of Technology in China.

Sad to say, suspicion falls on Yiyang Yao as the mastermind here. If, hypothetically, he had used his position as Reviewer and Guest Editor to run a rather amateur papermill (appropriating some manuscripts to repurpose them for his colleagues, and selling others to people in need of publications to adorn their CVs), the outcome would look very much like the situation we observe. Chao Xiong and his team would just be unusually regular customers.

In this regard, it kind of makes sense that Yao is a colleague of Caiyou Zhang at SGZEPC, while Luming Zhang is a collaborator of Caiyou Zhang. They three, together with Bo Dai, the first author of one of Caiyou Zhang’s 4 retracted papers from this special issue, have had close collaboration on more than one project. At least two of their collaborations ended up winning nominations for a National Award on Electrical Engineering Science and Technology, in 2017 and 2018. There is however no information on final results.

Based on available information on internet, Caiyou Zhang is in the high-level leadership position of the SGZEPC and apparently senior to Yiyang Yao or even Yao’s direct boss. In addition, Caiyou Zhang is the vice chief representative of Communist Party of China there.

What can be a reasonable speculation is that when Yao got the chance to organize a special issue (it could also be that the special issue was originally their idea proposed to the journal), he immediately thought of inviting his boss Caiyou Zhang to take advantage of the opportunity. Being not really in the academic field of the multimedia, Caiyou Zhang probably did not have much to submit for publication in that journal. Luckily, Yao was so nice to provide his own previously plagiarised recycled published paper for his boss to co-author for this new audience of the special issue.

 

Coda 1

Here and here are 2017 and 2019 Abstracts, by Yiyang Yao and two otherwise-disjoint teams of co-authors, paper-shredded for connoisseurs of the cut-up tradition… either that, or Google Translate was drunk that day. They are not yet unretracted

Image quality tendency modeling by fusing multiple visual cues

Yiyang Yao, Tengfei Wu & Jun Li

Image quality tendency modeling by fusing multiple visual cues

Wenjie Zhang, Yiyang Yao, Jinxion Wang, Xinyu Xiang & PengShu

 

If anything, the 2019 (JVCIR) instantiation of this Flickr-themed text is less grammatical than the 2017 version. It is like someone trained a neural network on the complete late works of Samuel Beckett. Perhaps it is deliberate, to veil the contents from the unknown original authors who might otherwise recognise their unpublished words in print.

“Flickr is a photo/video hosting site with over 87 million user. Upload more than 3 million 500 thousand new photos every day at present there are no tools to organize these huge numbers of users aesthetic tendency. Although Flickr allows users manually adding different groups, they are difficult to maintain Updates should be made when new users are added or deleted. In this paper puts forward a series of Flickr users system. Each loop contains similar users aesthetic tendency. We observed: (1) an aesthetic model of thought should be flexible, because of different visual features typically represent different data sets, and (2) Significant differences in the number of photos from different Flickr users stay. In this work, a new probabilistic topic model is proposed describe the aesthetic interest of each Flickr user potential spatial distribution. After that, an affinity graph is similarity is described by aesthetics interests of Flickr users. Obviously, intensive users of Flickr are similar in taste. Thus, these users are divided into different Flickr bounds efficient dense graph discovery. Piping it is proposed that the Flickr bound discovery is fully automatic. Extensive we show that our proposed method is accurate for mine Flickr experiments 60,000 Flickr user community.”

But wait! There’s more! For a third instantiation of the same Abstract had already found its way into MTAP, this time attached to different authorship and text. In fact it is the 14th recently-retracted entry, previously unmentioned, from that “Multi-source Weak Data Management using Big Data” Special Issue. That paper from 2017 also disappeared in May 2019, but not yet listed in Retraction Watch database.

RETRACTED ARTICLE: Flickr image quality evaluation by deeply fusing heterogeneous visual cues

Yongjun Zheng, Weiyu Di & Shen Jiang

“… because its content has been duplicated without permission from unpublished manuscripts authored by Sandeep Sood and by Pegah Nikbakht, Mojtaba Mahdavi, Shahram Etemadi and Sima Arasteh. The article also shows substantial overlap with a published article [2]. In addition, there is evidence suggesting authorship manipulation and an attempt to subvert the peer review process.”

Despite the seemingly-focused title, this third paper is longer, and globally incoherent; it was like the dreadful outcome of purloining three separate manuscripts and sending them though the matter transmitter simultaneously. Individual segments make a semblance of internal sense but in toto they are completely calenture frittiliary fishtank, lurching between (1) content analysis / aesthetic grouping of Flickr imagery, (2) digital watermarking algorithms for image validation, and (3) data encryption for secure data storage on the Cloud and the Internet of Things. There was handwaving and incantation of Worship Words but no concrete contribution to any of these concerns.

Some Figures re-appeared in the bilocating “image quality tendency modeling” paper(s) already adduced.

But though “Flickr image quality evaluation” is an unnatural hybrid which transgresses all the laws of literary lineage, it is also a kind of Ur-version, a textural trove to be plundered at leisure. So other Figures and many paragraphs of text found their way into a fourth publication, in MTAP again, retracted only days ago and not yet in Retraction Watch database

RETRACTED ARTICLE: Large-scale image-based fog detection based on cloud platform

Chao Xiong, Ruxi Xiang, Yuan Li, Xia Han & Hongwei Du

“… because its content has been duplicated from an unpublished manuscript authored by Sandeep Snood without permission. In addition, there is evidence suggesting authorship manipulation and an attempt to subvert the peer review process.”

You probably recognise the name of the ostensible author, which brings us back to where we started.

None of this makes any sense and now I have a headache.

Coda 2

It is all alarums and excursions at JVCIR, and even as I write the duplicated publications are vanishing from existence, in the manner of the memory-erasure scene from “Eternal Sunshine”. Perhaps the editors or managers read the Twiddle-box.

JVCIR has Special Issues of its own, and the papers briefly resurrected from MTAP all fell under the rubric of “Text-based Image/video Understanding in Social Media Context”. I have not found any record of the Guest Editors responsible for selecting and scrutinising the short-lived contributions.

---