Scopus has struck all links to the homepages of journals it indexes, Elsevier announced earlier this month. The move follows revelations that content from dozens of hijacked journals had been included in the database.
In a December 18 blog post, Scopus – which many universities and government agencies around the world use to create journal “whitelists” where authors are encouraged or even paid to publish – explains its rationale:
For example, scenarios may arise where the registration of a journal website lapses, or a genuine journal website is transferred to a new URL and the ‘old’ URL is taken over by another party. In a more severe situation, if a journal is hijacked, malicious actors may create a fraudulent duplicate of the journal’s website or mimic the appearance of the original journal to deceive researchers.
Retraction Watch readers may be familiar with the phenomenon of hijacked journals, which we have cataloged since 2022. In November, I published a paper showing that data from 67 legitimate but hijacked journals had been compromised in Scopus in the decade leading up to September 2023.
Among these, 33 indexed unauthorized content, meaning papers published on a fake website were covered by Scopus. Twenty-three hijacked journals compromised the homepage link of a legitimate journal, redirecting it to a hijacked, rather than a genuine, journal. In 11 cases, hijacked journals succeeded in compromising both the content and the homepage link.
Elsevier told Science in a story reporting on that paper on December 5 it had begun an investigation of the issue. The publisher announced on December 18 it planned to remove journal URLs.
The indexing of hijacked journal papers continues, however. A day after the December 18 blog post, Scopus alerted some users, including me, it had indexed new documents from the Hunan Daxue Xuebao/Journal of Hunan University Natural Sciences. The email referred to a paper titled ‘Solution to the Klein-Gordon Equation Using FEM,’ published in issue 50 (7) of Hunan Daxue Xuebao/Journal of Hunan University Natural Sciences on pages 233-240.
The problem is that Scopus has already indexed a paper from the same journal, issue, and overlapping pages, as demonstrated from this screenshot taken December 19:
This paper ‘Solution to the Klein-Gordon Equation Using FEM’, along with many others, was published on the website jonuns.com, which hosts a cloned version of Hunan Daxue Xuebao/Journal of Hunan University Natural Sciences. Scopus currently indexes papers from both legitimate and cloned journals that result in different paper titles published on overlapping pages of the same issues of the journal.
This journal is one of the 67 identified in my November paper, which documented entries until September of this year.
Since September 2023, at least three hijacked journals had their links on the homepages of legitimate journals, redirecting to fraudulent sites (the links in these screenshots have since been deleted):
*
How these hijacked journals managed to change the link rather quickly is unclear, especially when honest journals, such as the Scandinavian Journal of Information Systems, struggled for many months to reclaim their links back to the profile of the journal.
Elsevier notes in its blog post that Scopus now allows the homepage to be accessed through a DOI, for which the publisher is responsible:
Since the “View at Publisher” links from Scopus to the full text of the article will still be available, it may be possible for a user to reach the journal homepage via the full text link. Unlike the Source Homepage link, the “View at Publisher” link to the full text is created through the DOI of the article and the destination of that link is managed by the publisher itself. The quality of these links is therefore not a concern.
However, the ‘view at publisher’ link for the Hunan Daxue Xuebao/Journal of Hunan University Natural Sciences paper indexed in Scopus earlier this month led directly to a hijacked journal. That suggests that deleting homepage URLs will not solve the problem of hijacked journals infiltrating Scopus, as non-authentic content has been present in the database since 2013.
(The paper has since been deleted from Scopus, but some other unauthorized content in this journal remains.)
Of note, many journals indexed by Scopus lack the ‘view at publisher’ option, either because they do not have DOIs or because they are published only in print. Furthermore, the ‘view at publisher’ option is only available in the subscription version of Scopus.