I have a large .bib file and need to analyze which keywords are used in which bibliography entry. Therefore I am looking for a "tool" to convert the .bib file to CSV or XLS format.

I need a CSV which have information about:

  • CitationID (=Name of the biblatex entry)
  • Title
  • Keywords.

3 Answers 3


This is very easy:

  1. Install JabRef. You find the download page here. JabRef is free software.
  2. Load your bib-file into JabRef, i.e. Choose “Open database” from the file menu.
  3. For security reasons, save the database under a new name (File; Save As) so you work on a copy (better safe than sorry if you do something wrong)
  4. Open the file menu, choose “Export” from the options, and choose the file format you want to export to. Excel 2007 is one of the options.
  5. If you need to clean up or manipulate the resulting file, export to CSV-files and use Sam Franche’s CSVed or uniCSVed if you are on Windows.
  • As of Nov 2018, I can't see an option to export in Excel format from JabRef?
    – seinecle
    Nov 14, 2018 at 9:31
  • Does the first way really work? For example, I have multiple entries in the excel file for publications with multiple authors (one for each author)...
    – Alex
    Mar 28, 2019 at 14:57

You do not need at all export the .bib file to .csv or .xls format.

It is possible print references by keywords with LaTeX (cited or not in the text) using biblatex.

Example: Suppose that all references have fields as keywords = {XX} or keywords = {XXI} to distinguish what was published in each century. Then you want to show only the cited papers of the XX century.

Bla bla \cite{Guy1998} and \cite{Guy2012}

This should print only the reference of Guy1998. To show also the not cited references of the XX century, add \nocite{*} before of \printbibliography. You can use this command several times.

Remember that by default you should compile the biblatex bibliography with biber(not with bibtex), after compiling with pdflatex and then compile with pdlatex at least two times more in complex documents to solve all the cross-references. No so easy, but with biblatex you can also ignore some fields of the references, if that matter.


There are at least 4 ways.

  1. Using JabRef. You can follow Sveinung's answer.
  2. Using LaTeX. As explained by Fran.
  3. Using Zotero. The poin is to read the data inside the Zotero SQL and save it as csv. Read the procedure had been prepared by Roy(2018) here: http://roycekimmons.com/tutorials/zotero_to_excel
  4. Using Direct Citation (References). The idea is to read the reference directly to excel based on the available delimiter. The example of using MLA reference format had been writted by Brennan(2016) here: http://doi.org/10.3163/1536-5050.104.1.012
  • Welcome to TeX.SE! Could you explain how?
    – CarLaTeX
    Aug 24, 2018 at 15:24
  • Thanks. I had add all references. Point 3 and 4 had not been coped in the Sveinung's and Fran's answer. I try my best to keep the answer brief so that the reader can read fast. Please give me more advice if I still need to do something else. I do apreciate it. Aug 24, 2018 at 19:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.