This is a report on the problem with my Google Scholar Citations. More than 30,000 citations have been lost due to some bug(s) in the Google Scholar algorithms, and it took about one year to get it fixed and corrected. Fortunately, the Google Scholar team paid attention to my online updates. Many thanks to them for that!
Please read and double-check your own Google Scholar profile, or continue to my regular "index.html".
Finally, today it looks correct! It took one year... But now Google Scholar sees my "Fractional Differential Equations" book, and correctly attributes its citations:
And now my total citations count also looks correctly:
So, happy end? Maybe. Who knows? When big data are processed. a small bug might cause a notible error.
It becomes really annoying: the Google Scholar still does not see my book. The following video is practically the same as the video that I posted here on August 28, 2022 (compare both videos yourself):
Dear Google Scholar team, here are the links to my book at Elsevier (the publisher), your Google Books, and Amazon.
Although it is already ONE YEAR of this persisting issue or bug in your algorithms, I still hope that you can fix the problem with the (in)correct indexing of my book and its citations.
This looks like a never-ending story! Today, March 21, 2023, I have noticed that my total citations counts return to normal numbers (53,369 citations), but again in a wrong way — citations to my FDE book are again attributed to someone else!
Although that book item titled "Fractional differential equations" (sic! people can think that it is my FDE book...) by Zhi-Zhong Sun and Guang-Hua Gao exists (and has so far only 13 citations according to the book page at the publisher...), this is still an obvious error on the Google Scholar side (or in its algorithms). Google Scholar simply mixes and does not distinguish my book and the SUN-GAO book; clearly, Google Scholar clustering or aggregation algorithms still obviously contain a bug (or bugs).
Clicking on the number of citations "33994" in the previous screen and limiting citing publications to the period of 1999-2019 produced the the screenshot below. However, the Z.-Z. Sun's and G.-H. Gao's book was published in 2020, and it cannot receive "about 25,000 citations" before its publication (exactly like it was in the case of the SAR-GIRESUNLU paper, which I described earlier)... So, those 33994 citations are actually citations to my FDE book.
I have to add that the SUN-GAO book does not appear to be indexed in the Web of Science; it is indexed only in Scopus, which shows for it only 14 citations:
I wonder when (if at all...) the Google Scholar team will fix the error in their indexing and clustering/aggregation algorithms. Perhaps sometimes it is better to use natural human intelligence instead of relying on the AI?..
Two side remarks:
At 18:25, Jan 26, I have noticed that my citations dropped down again, from 51K+ to 19105 citations(see first screenshot below). I can only hope and believe that the Google Scholar team is working on fixing the problem — because wrong citations count totally disappeared from one of the instances of the SAR-GIRESUNLU paper (see the second screenshot below).
Note that now there is nothing (no citations count) near the second instance, and even more — now it has the correctly labelled link (not "Fractional Differential Equations", like before). Here I observe some partial fix.
However, the Google Scholar still does not see proper records of my book (very strange; it is available at the Elsevier website, at Google Books, and in many other web locations); the Google Scholar sees only something that it denoted as [CITATION]:
So, compared to the state of 2022-12-03, there are changes. The good one is that the citations to the SAR-GIRESUNLU paper have been corrected; this is definitely the first necessary step. The bad one is that my citation count dropped down. Perhaps I have to wait (a week? a month? or how long?) until the Google Scholar algorithms will bring this to correct numbers of citations to my book...
During the weekend of November 19-20, 2022, I have noticed that the citations to my book re-appeared on my second profile, which I used for investigating the problem with disappeared citations. So, I had to hide my original profile with the ID 4H2CInEAAAAJ, and make the second (auxiliary) profile with the ID QQkfeNkAAAAJ visible.
But — I have noticed that there is still an obvious problem: my correct 30,000+ citations are linked to a wrong article by other researchers, and that article, in fact, has only 17 citations in the same Google Scholar... The details are below.
Yes, there should be a bug, or damage, in the Google Scholar dataset. Here is what I have found when I was investigating the mistery of the disappeared and reappeared citations to my book.
(1) The total number of citations of my book returned to normal value:
(2) However, when I click on the title, I see that Google Scholar added something strange -- see the first item on the next screenshot. I did not add that item (I did not merge it with other items which are correct here). It is a short (only seven pages) article by two Turkish researchers (Emrullah Yaşar and Ilker Burak Giresunlu) whose names were unknown to me until now, and the Google Scholar shows 32883 citations to it! It is denoted as "Fractional Differential Equations *", and the star symbol normally means (in Google Scholar Citations) that this item has been edited manually. Moreover, the name of Emrullah Yaşar appears in the wrong form as: EYA SAR. Below I use the short name SAR-GIRESUNLU for this article.
Let us take a look at those 32833 citations -- see the video right below this paragraph. We first search in Google Scholar for all works having the string "fractional differential equations" in their title, and then go year by year from 1999 to 2022 examining how those 32883 citations aggregated with time. Clearly, the SAR-GIRESUNLU article, that has been published in 2016, could not receive "about 12,500" citations during the period of 1999-2015 preceeding its publication! Those 32833 citations are citations to my FDE book published in 1999.
So, now it is clear that citations are citations of my book, but the description of the item provided by Google Scholar is wrong... What is the cause? Let us look further.
(3) Note that there is the link denoted "All 6 versions". Clicking on that link shows the following cluster:
(4) Let us try to find out if Google Scholar sees my book. For this, we should use the "allintitle" parameter, which finds all documents containing the given string in their title. My book does not appear in the list, but this strange item does... with the citations belonging to my book...
(5) Now let us take a look at that strange item. It is a 7-page paper in Pramana - J. Phys. published in 2016. We see that it contains the string "fractional differential equations" in the title, but my name does not appear in that paper – so there is absolutely no reason for linking the SAR-GIRESUNLU article to my name.
(6) Let us find the real citations count for that article. Using its long title for search in Google Scholar gives two outputs: one of them has its proper long title and 17 citations, and this is the correct number; the other one strangely has the title of my book and 32883 citations – and this is wrong, because those citations are citations to my book.
(7) The search for the SAR-GIRESUNLU article in the Web of Science shows that it is definitely not a highly cited article, because it has only 12 citations there, which confirms that it cannot have those 32883 citations (of my book):
(8) One of the authors of that article has his Google Scholar Citations profile (I.B. GIRESUNLU), and on his profile that article has the correct number of citations (17 citations):
(9) In the metatags of the the PDF file of the SAR-GIRESUNLU article I cannot see any information looking like "fractional differential equations"; my name does not appear in that article, too. It is not clear why should the Google Scholar label the SAR-GIRESUNLU article as "Factional Differential Equations". It is not also clear why two unrelated names were added by the Google Scholar to the title of my book and to its citations records.
Conclusion:
I believe that it could be a damaged dataset, or it could be a wrong manual correction of the data by some person having access to the dataset.
I would like to repeat what I wrote previously. The probability of my book being the one-and-only affected publication is zero (however, remember that zero probability of an event does not mean that the event is impossible). Finding an error and fixing it might most probably help other researchers, too. The Google Scholar plays already an extremely important role in the scientific community, for doing research and for evaluating research performance, and it must work as reliably as possible.
Igor Podlubny,
December 03, 2022.
During some recent update of its service, the Google Scholar Citations, Google lost about 30,000 citations to my highly cited book "Fractional Differential Equations" (Academic Press, Inc., San Diego, 1999).
I understand that the amount of data that Google Scholar collects and processes is enormous, but it might be that I am not the only one affected by the recent update.
My Google Scholar Citations ID is: 4H2CInEAAAAJ, and my profile is: http://scholar.google.com/citations?user=4H2CInEAAAAJ
The severely affected publication is my book (further "FDE"):
Podlubny, I.: Fractional Differential Equations. An Introduction to Fractional Derivatives, Fractional Differential Equations, to Methods of Their Solution and Some of Their Applications. Academic Press, San Diego-Boston-New York-London-Tokyo-Toronto, 1999, 368 pages, ISBN-10: 0125588402, ISBN-13: 9780125588409, eBook ISBN: 9780080531984.
(https://www.elsevier.com/books/fractional-differential-equations/podlubny/978-0-12-558840-9)
~
On April 22, I had a meeting with the deputy director of the Isaac Newton Institute (Cambridge, UK), and on that day my FDE book had 31,761 citations.
However, two months later, on June 20, the Google Scholar showed only 2,419 citations to my FDE book -- less by almost 30K lost citations (29,342 citations have been lost, to be precise)... And on August 28, 2022, the Google Scholar still shows only 2547 citations to this book... !
This is obviously wrong, because:
To find citations to my book in the Web of Science, the "CITED REFERENCES SEARCH" must be used. On August 8, 2022, there were 20,497 references to my book in the papers indexed by the Web of Science. See the video below.
In the end of the video recording you can see that, according to the Web of Science analysis, my FDE book is cited in 133 different fields of science, and the number of its citations only in the single year 2021 is 2,497...
In Scopus, the "Search within [References]" must be used. On August 8, 2022, there were 22,083 references to my book in the papers indexed by Scopus. See the video below.
According to MathSciNet (Mathematical Reviews, online) database maintained by the American Mathematical Society, my book "Fractional Differential Equations" currently has 5,320 citations in the works indexed by MathSciNet, and is the most cited mathematical book of the decade 2012-2021, being #1 in 2012-2014 and 2017-2021, #2 in 2016, #4 in 2015.
Even those 5,320 citations indexed in MathSciNet with its narrow mathematics-only scope is two times more than the number of citations of my book shown by the Google Scholar on August 28, 2022.
After noticing that a huge number of citations to my book disappeared from my Google Scholar Citations profile, I tried to see if Internet Archive (web.archive.org) eventually has any archived copies of my profile. And it had some copies; the latest is dated March 5, 2022 (strangely, in Swedish language), with 31,113 citations to my book.
Go to Google Scholar, and search for [ podlubny "fractional differential equations" ], as shown in the video below. This will return only one strange item...
My hypothesis is the following. As far as I understand, the Google Scholar more or less periodically performs updates of its dataset, and then runs its algorithms for processing the publications and citations on that updated dataset. Most probably there is an error at the stage of the dataset update.
The probability of my book being the one-and-only affected publication is zero (however, remember that zero probability of an event does not mean that the event is impossible). Finding an error and fixing it might most probably help other researchers, too.
I hope that it will be possible to fix the problem with citations to my FDE book soon during the update of the otherwise excellent Google Scholar Citations product. The Google Scholar plays already an extremely important role in the scientific community, for doing research and for evaluating research performance, and it must work as reliably as possible.
Igor Podlubny,
August 28, 2022.