18

After having a paper published, I submitted the preprint to arXiv, but with a different title.

Now, Google Scholar has identified the arXiv version as another paper, and by merging the two articles using the merge button, I cannot make the arXiv version appear next to the published paper when someone searches for it.

I guess that Google looks for the title of a paper (in the PDF, not the file name), and decides if this is the same paper. Anyway, is there any way to make Google Scholar understand that the arXiv version is the same paper and show the PDF in the search results?

Improve this question
6
  • 2
    Why did you change the title? Is it just to clarify it's a "preprint"?
    – Fuhrmanator
    Nov 16, 2014 at 18:56
  • 16
    I submitted the preprint to arXiv, but with a different title — Don't do that.
    – JeffE
    Nov 17, 2014 at 0:01
  • 2
    Is there a specific reason why you require the arXiv preprint to have a different title to the published version? Ideally after publication you should update the arXiv to be as close as possible (including the title) to the published version. Journals often change the title regardless of the authors' wishes, but making two versions of the same piece with different titles will not only confuse Google, it will confuse every reader that sees both and is forced to disentangle their relationship.
    – E.P.
    Nov 17, 2014 at 2:00
  • 2
    How do you plan to convince all the researchers in your field that it's the same paper? How many of them are going to think, "Oooh, a new paper by user24369 et al. That looks interesting... Damnit! I already read it. Thanks for wasting my time!"
    – David Richerby
    Nov 17, 2014 at 10:36
  • 3
    @E.P. I completely agree that this is confusing. But the opposite can be equally confusing. I know one case where a paper exists in three different versions all with the same title and authors: an early arXiv, which apparently never made it to publication; the published, significantly changed version; and the “actual preprint” corresponding roughly to the published version.
    – xebtl
    Nov 17, 2014 at 11:50

2 Answers 2

Reset to default
26

This may not be the answer you're looking for, but one solution would be to update the arXiv paper so that its title agrees with the published paper. (You can change an arXiv paper's title by submitting a revision.) Maybe you prefer the title from the arXiv, but using different titles causes enough hassle and confusion that I can't really believe it's worthwhile.

It's not just a matter of convincing Google Scholar. If they supply a link to a paper with a different title, some users will assume it's an error without looking closely enough to detect that it's really the same paper. The same issue will occur whenever anyone runs across the arXiv paper, since the first heuristic most people use to decide whether two papers are the same is comparing the title and authors.

If you really want to use a different title, you should take every opportunity to clarify the relationship between the papers. For example, the arXiv abstract page and the first page of the article should explicitly state that it's the same as the published paper (and give the citation). But even if you do that, readers will be confused and perhaps a little annoyed. They'll naturally wonder why it has a different title if it's the same paper otherwise, and they may wonder what else you have changed compared with the published version. If there are nontrivial changes, then you should warn the reader, while if there aren't, then it's not good to let readers wonder about that.

Improve this answer
1
  • +1 for updating arXiv, that is what I wanted to suggest. It would be interesting for @user24369 to explain why they changed the title in the first place.
    – xebtl
    Nov 17, 2014 at 11:47
8

In my experience, when you use 'merge', Scholar will not return multiple listings in the main result. Instead, it asks you which version is the 'better' version and shows that version preferentially. If a person clicks on the 'all N versions' button, however, the alternate version should appear in that list.

If you want to have a more explicit statement of the two articles and their relationship, you cannot force the search engine to do it for you---and even if you could, it wouldn't necessarily remain that way next year, since Google is always tweaking their systems. Instead, you should put this on your personal webpage, which (if hosted by your institution) will likely end up high in Scholar's returns in any case.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .