A study on a wireless communication algorithm was retracted for being an exact duplicate of a paper submitted to a separate journal last year — but the authors were different and it’s unclear how they got hold of it.
The retracted study, “Energy-aware resource management for uplink non-orthogonal multiple access: Multi-agent deep reinforcement learning” was published in the Elsevier journal Future Generation Computer Systems. Neither the author of the original work who we were able to reach, nor either journal involved, say they know how the unpublished manuscript got into another group’s hands.
Here’s the (complicated) timeline:
On June 27, 2019, a group of researchers submitted a conference paper to the 2019 11th International Conference on Wireless Communications and Signal Processing (WCSP) — to occur in October.
Seven weeks later (August 18, 2019), the authors submitted a fleshed out version of the paper to the journal IEEE Access, only to have it rejected 12 days later.
Two days after that, a different group of authors submitted an identical manuscript to the Elsevier journal Future Generation Computer Systems, which accepted it on December 29 and published it on January 2, 2020.
Meanwhile, the authors’ conference paper had gone live on IEEE Xplore December 9, 2019.
It was at that point that Xiaoming Wang of Southeast University in Nanjing, China, and one of the authors of the WCSP conference paper, noticed something odd: Wang told Retraction Watch in email that the duplicate manuscript, by Yingfang Li and co-authors, also copied the original’s citations.
Wang wrote:
We found the duplicate article when searching the list of citing our published papers on Google Scholar, and then we contacted the journal.
On June 15, 2020, Elsevier retracted the Future Generation Computer Systems version.
The retraction notice says:
This article has been retracted at the request of the Editor-in-Chief, authors and Xiaoming Wang.
This article is the identical resubmission of a paper that was initially submitted by [a] different group of authors, prior to this publication, to IEEE Access in August 18th and was rejected on August 30th. A conference version of it was published in 2019 11th International Conference on Wireless Communications and Signal Processing (WCSP)…. There is no evidence of the authors of this article having access to the IEEE Access submission during peer review.
Apologies are offered to readers of the journal that this was not detected during the submission process.
Wang told Retraction Watch:
Our paper rejected from IEEE Access is an expanded long version on WCSP’19. Yingfang Li’s paper is a full-text copy (except for title) from our manuscript submitted to IEEE ACCESS.
The WCSP paper was called “Energy-Efficient Resource Allocation in Uplink NOMA Systems with Deep Reinforcement Learning.” The other two authors were based out of Nanjing University in China.
The editor of IEEE Access, Derek Abbott, forwarded our request for comment to IEEE. A spokesperson for IEEE, responding in email, declined to comment on how the authors might have accessed the Access manuscript or why it was rejected.
Peter Sloot, the editor in chief of Elsevier’s Future Generation Computer Systems, forwarded our request for comment to Elsevier, whose communications director, Andrew Davis, told us:
IEEE Access confirmed the submission and rejection of “Energy-Efficient Resource Allocation with DRL for Uplink NOMA Systems” by Xiaoming Wang and found no indication that the IEEE Access reviewers would have a connection to the Elsevier manuscript…
As for the original manuscript, he wrote:
There were no other issues with the paper. IEEE ACCESS did not specify the reasons for the rejection of the paper from the journal.
He added that because the first paper was never published online, the journal’s plagiarism checks weren’t able to detect the duplication.
Li and the other authors of the duplicate study, from Honghe University, in Mengzi, China and Yunnan Normal University, in Kunming, China, did not respond to multiple email requests for comment. We could not find contact information for Bo Yang.
Wang wrote:
We didn’t show the IEEE Access version paper anywhere else and we didn’t know Li et al before. But IEEE Access journal have confirmed that all of them (Yingfang Li et al) are not the reviewers or [associate editor].
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
Perhaps some editorial service or even a paper mill had worked on the first paper before its submission and then sold the text of the same paper to the second group of authors. Almost any form of review of a paper before submission can raise the possibility that the paper gets leaked.
A similar albeit less obvious occurrence is described here https://www.timeshighereducation.com/features/each-scientist-must-stand-all-costs-truth
except that the journals involved did not take any action.
Some of the mystery behind the plagiarism of an unpublished paper outlined in this article can perhaps be explained using the following information:
Wei Gao, one of the authors of the retracted paper has already published 5 papers with IEEE Access in the calendar year 2019 alone. So, the authors of the retracted paper were well versed with IEEE Access submission and publication procedures. There is more information I have gathered about some of the dynamics as to how this may have happened, but I will verify this in depth before commenting further on the rest of the mystery.
Anyway, I found that another paper authored by some of the same authors retracted as well. The details of the paper are as follows:
Title: A Fast Implementation of Bilateral Filter Based on Edge Protection
Authors: Li Ying Jiang, Yan Li and Yang Bo
Journal : The Open Automation and Control Systems Journal, 2015, 7, 275-283
Link to paper: https://benthamopen.com/contents/pdf/TOAUTOCJ/TOAUTOCJ-7-275.pdf
Link to retraction notice: https://benthamopen.com/ABSTRACT/TOAUTOCJ-8-104
Interesting, if maybe not probative, is that part of this mystery came wrapped in an Elsevier journal.
I am the Editor-in-Chief for the Taylor & Francis journal, Tourism Geographies. We had an identical situation come up a couple of years ago. The paper that I received had data for several cities in Turkey. It was sent for review and one of the reviewers reported that it was the exact same paper that he had written and which was recently accepted by the journal, Tourism Management (Elsevier) — except that the city names in his paper were all in South Korea. The Korean version of the paper was still in production and had not yet been published. Somehow, the Turkish authors had gotten hold of this reviewer’s original paper and simply swapped the Korean cities for Turkish cities, and then tweaked the text accordingly.
We do not know how this happened. I informed the EIC for Tourism Management about this, but never heard if there was any follow-up. If this paper had not been sent to the author of the original paper, it *might* have eventually been published in my journal, and eventually retracted.
In continuation with my previous reply to this post, I found an interesting connection to this story:
Derek Abbott (Professor with the School of Electrical and Electronic Engineering), the editor-in-chief of IEEE Access, who forwarded RetractionWatch’s email to IEEE, published 8 papers in IEEE Access in this calendar year alone (in just under 8 months time). Derek Abbott was editor-in-chief of IEEE Access for all of this time.
The list of these papers are as follows:
(1) Exploring Low Loss and Single Mode in Antiresonant Tube Lattice Terahertz Fibers
(2) Quantification of Atherosclerotic Plaque Elasticity Using Ultrasonic Texture Matching
(3) Experimental Study on Glass and Polymers: Determining the Optimal Material for Potential Use in Terahertz Technology
(4) Universal Kriging Prediction of Line-of-Sight Microwave Fading
(5) Tunable Terahertz Graphene-Based Absorber Design Method Based on a Circuit Model Approach
(6) Angiogenic Networks in TumorsInsights via Mathematical Modeling
(7) Exploring Bona Fide Optimal Noise for Bayesian Parameter Estimation
(8) Comparative Study of Discrete PI and PR Controls for Single-Phase UPS Inverter