(N.B.: The following was created as a Markdown file in _posts/. For Jupyter notebooks, the same things apply; one simply enters the Liquid codes into Markdown cells. Quick Jupyter example.)

How to Cite

For demonstration purposes, I’ll take the liberty of citing a couple of my recent papers, namely the first SignalTrain paper [1] and the new one by Billy Mitchell [2]. Instead of using the LaTeX code \cite{ <whatever> }, I use the Liquid code {% cite <whatever> %}. For example, the first citation above is written as “{% cite signaltrain %}” in the Markdown file that generates this HTML page.

The two citation markings above point to the References section at the end of this post where the full references are printed out in the bibliography style of my choice.

Drawing from the Bibliography

In the main blog directory, create a new directory called _bibliography/, and place your BibTeX file there as references.bib. In the case of this demo, the references file looks like this:

@conference{signaltrain,
title = {Profiling Audio Compressors with Deep Neural Networks},
author = {Hawley, Scott H. and Colburn, Benjamin and Mimilakis, Stylianos Ioannis},
booktitle = {Audio Engineering Society Convention 147},
month = {Oct},
year = {2019},
url = {http://www.aes.org/e-lib/browse.cfm?elib=20595}
}

@article{billy_signaltrain2,
title={Exploring Quality and Generalizability in Parameterized Neural Audio Effects},
author={William Mitchell and Scott H. Hawley},
journal={ArXiv},
year={2020},
volume={abs/2006.05584}
url = {https://arxiv.org/abs/2006.05584}
}


Note that this (single) references file is for your entire blog. The great thing about this is that all your Jupyter notebooks and Markdown posts will draw from this same file, which could be hundreds of references long, and jekyll-scholar will only include the ones you need for each post.

Finally, at the end of your post, you signal the creation of the list of references by using the Liquid tag

{% bibliography --cited %}


…so I’ll put that at the very bottom of this file. (Currently that’ll generate an error, because we haven’t enabled jekyll-scholar yet, but we’ll do that next.) The optional argument --cited means it’ll only list the references cited in your post.

Enabling Jekyll-Scholar

To enable jekyll-scholar, all we need to do is make the following two changes, and perhaps a third.

1. In _config.yml, add “ - jekyll-scholar” to the list of plugins:.

2. Edit the Gemfile to include gem 'jekyll-scholar' where the other plugins are listed.

3. Optional: The default citation format is “apa”. If you want to change that, you can add the following to your _config.yml file:

scholar:
style: <name>


…naming one of the styles in the CSL style repository (but leaving off the .csl ending). Tip from the CSL maintainers:

To quickly search the styles in the GitHub CSL style repository by file name, press “t” to activate GitHub’s File Finder and start typing.

Note however that the csl-styles Gem package used by jekyll-scholar lags behind the official CSL style repository, so some names you choose may not work. In that case, you can supply a CSL file yourself. For this demo, I found the file physical-review-d.csl, added it to my main blog directory, and then specified the style name physical-review-d in _config.yml. This produced the bracketed-number citation markers above, and the reference format you see below in the References section. (EDIT: Actually I customized the CSL file a bit after that, so my new file is custom.csl.)

The convenience of this BibTeX/jekyll-scholar approach is that instead of having to manually edit references on each individual page – say, if you wanted to change citation formats (or alternatively, update information about a paper cited in multiple posts) – now you only change one line in _config.yml (or update one spot in references.bib) and the system “builds out” the change “everywhere.”

Happy blogging!

References

[1] S. H. Hawley, B. Colburn, & S. I. Mimilakis, "Profiling Audio Compressors with Deep Neural Networks," Audio Engineering Society Convention 147 (2019).

[2] W. Mitchell & S. H. Hawley, "Exploring Quality and Generalizability in Parameterized Neural Audio Effects," ArXiv abs/2006.05584 (2020).

(c) 2020 Scott H. Hawley