I am reading a review paper which includes hundreds of useful citations for my paper. I would like to import some of the citations into a bibtex file. The only thing I know to do is to search them on google one by one, and then click on "export bibtex".

Is it acceptable (ethical) to ask the author of a paper for the original bibtex (Latex) file?

Or I wonder if there are any program could do the repetitive job: searching each citation on google one by one and export the bibtex from google scholar.

Thanks for the Scopus answer. It works for most of the published (not working papers, though) papers!!

  • 20
    If you just ask for the bibtex file rather than the rest of the latex source, it might be seen more positively. On the other hand, some people use one giant bibtex file for all their projects, and might not like to share with you the list of every paper they have ever looked at. Commented Sep 30, 2017 at 22:47
  • 24
    It would be kind of weird, though I probably would share it if asked. FYI: If the paper is posted to the arXiv, you can download the source from there without asking the author. Commented Oct 1, 2017 at 1:29
  • 5
    You mean .bbl? That is only useful if you want to use exactly the same citation style as me. (By the way, I think arXiv only needs the .bbl as well, so @AndyPutman's suggestion may suffer from the same issue.) Commented Oct 1, 2017 at 3:05
  • 5
    The title of this question currently says "original Latex file", which makes it sound like you're asking for the source of the whole paper. That is a rather different request than source for the bibliography, and I would recommend editing your question to clarify the title.
    – Mark S.
    Commented Oct 1, 2017 at 4:01
  • 14
    This has nothing whatsoever to do with ethics. Commented Oct 1, 2017 at 12:32

5 Answers 5


I would consider this a perfectly reasonable, if unusual, request.

That probably has a lot to do with the culture of my field; most papers in particle physics are uploaded to arXiv as LaTeX source (though .bib files are often not included), which is then made available to download along with the rendered PDF. This facilitates a cultural expectation that source code to the paper should be available to readers, not hidden. So I would consider sharing the BibTeX file associated with a paper to be a reasonable thing to do, a professional courtesy of sorts - similar to what sharing a copy of the paper itself is or used to be.

Other fields may think differently. Still, in most cases I would think it's fine to ask. The author can always say no or ignore your request, if they don't want their file to be shared for some reason.

  • 6
    This. I actually do make sources of my papers freely available on my website, .tex and .bib files both. I honestly have no idea why anyone would make a fuss about it. I sometimes do look at .tex files of other people's papers (on arXiv) to see how they typeset something. Sure, sometimes I found slightly embarrassing hacks, but it's not such a big deal. ;-)
    – tomasz
    Commented Oct 1, 2017 at 9:58
  • 1
    @ElizabethHenning: my point is, I don't see what the big deal is about making it publicly available in the first place. That or sending a copy to someone. This is of course different if you don't have access to the source for whatever reasons, but otherwise...
    – tomasz
    Commented Oct 1, 2017 at 18:26
  • 6
    @Elizabeth FWIW the .bib files are often not included in the source uploaded to arXiv.
    – David Z
    Commented Oct 1, 2017 at 18:30
  • 2
    @Gabriel - sure, unless it appears to take some time to prepare the bib file for it to be 'ready' for random consumption. (remove extraneous items, or at least check and get rid of any 'review' or 'note' comments associated with entries that might not be notations that one would want distribute.
    – Carol
    Commented Oct 2, 2017 at 17:32
  • 1
    @Carol I've never added any comments or notes to my .bib files, but if that's the case removing them can not be such a time consuming task. If it indeed is, then sure, you are not obligated to take the time.
    – Gabriel
    Commented Oct 2, 2017 at 18:35

My feeling about this is that if you don't personally know the author, it's weird and tacky and will probably result in being ignored. Even if you do personally know the author, it's still kind of weird and tacky. But there's nothing unethical about it.

  • (upvote) I think it is a little bit awkward, too. Is it better to mention this in a friendly, personal chat (if this is an option)?
    – High GPA
    Commented Sep 30, 2017 at 22:54
  • 18
    @HighGPA You're basically asking someone to (1) share their working documents with you for the purpose of (2) saving you some work which is really your own responsibility. I'm not saying you should categorically not do it, but it could be seen as very presumptuous depending on your relationship with the author. Commented Sep 30, 2017 at 23:06
  • 8
    "weird and tacky"? I can honestly not think of a single reason why anyone would consider sharing the sources of a published article (which are already public) either "weird" or "tacky"... I guess this must be related to the different disciplines of research.
    – Gabriel
    Commented Oct 2, 2017 at 16:10
  • I disagree. Reusing a bibtex file would be an example of working efficiently, which is a good thing. Commented Nov 16, 2020 at 3:17

As an alternative solution, if your institution has access to Scopus, you can find the paper there and see the citation list and export it in Bibtex. Web of Science has a similar feature.

To answer your question, however, it is not really unethical, but would likely be seen as a weird request.

  • (upvote) My current school only gives the access to Scopus's competitor, Web of Sci. Does Web of Sci has similar function?
    – High GPA
    Commented Sep 30, 2017 at 22:53
  • 3
    Yes, you can find the paper and click on "Cited References". That will give you the bibliography and allow you to export them. You can't do it directly as a Bibtex, but you can output to a couple of bibliography management software, or output the raw citation data which isn't difficult to convert to bibtex. Commented Sep 30, 2017 at 23:01
  • Thanks! Luckily I have the access to Scopus through my another affiliation. The only problem is the most useful review paper is most likely a working paper which is not documented by Scopus......
    – High GPA
    Commented Sep 30, 2017 at 23:11
  • 9
    The .bib files produced by Scopus/WoS/Google Scholar are not directly usable without manual editing. They get a lot of things wrong (capitalization, name/surname splitting, special characters...). Commented Oct 1, 2017 at 12:05
  • 2
    @JAB I am not sure of what you are trying to tell me with this comment. Do we agree that grabbing someone else's bibtex file (with a useful selection of references, human-readable keys, and some manual edits to fix at least glaring mistakes such as those that cause compilation errors) requires a lot less work than trying to produce a new one from automated sources from the internet? Commented Oct 1, 2017 at 21:13

The context can often make a big difference. If you are an experienced researcher, then this request would be seen as weird and possibly annoying since you could easily find these .bib sources yourself without bothering busy people. However, if you make it clear in your request that you are a high school student without proper [experience and] resources, and also make clear that you don't want them to go to any bother to accommodate your request, then I think that most people would be happy to help, and it wouldn't be awkward.

  • 1
    I know how to produce the bib source by myself. But I am not sure if it is "easy" to find the bib source by myself?
    – High GPA
    Commented Oct 1, 2017 at 16:40
  • In your situation, where you might not have proper resources [or experience], I would not call it "easy" for you - so it would fine for you to email a request for help. On the other hand, if you were a professional researcher, then you would have resources and experience that would allow you to create the bib source yourself "easily", so an email request would be far less appropriate. Commented Oct 2, 2017 at 23:13

There is the Arxiv API that can be used for this purpose.

You can read the documentation and programmatically make requests for the papers that you need, asking for the various fields as an output. You can send these requests as you want from the browser or any other programming language. In python for instance you can use urllib.request.

Furthermore, someone already did this and it is freely available on github.

For physicists there is an alternative using the INSPIRE API that is documented here. It can be used to automatically request for the bibtex entry without the need of composing it by hand by reading the various fields.

  • No, I should have explained better. An API is a service that websites offer you to programmatically request data from them (with some limitations of course). This means that you can write an URL that says "Arxiv, give me the bibtex citation data for the papers with id 2011.12345, 2011.12346, ...". Which will return either an html page, a json file or a text file with the data requested. And then you can make a python script that generates said url automatically. Commented Nov 16, 2020 at 21:27
  • Truly sorry about misunderstanding you! Does you method only works for sites like Arxiv or for any journal articles?
    – High GPA
    Commented Nov 17, 2020 at 0:44

You must log in to answer this question.

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