Contributing articles

If you are a student at the Hertie School who is interested in sharing your project, please follow the instructions below to contribute an article to this blog.

General Instructions

Articles posted on the Data Science Student Projects website are authored using Distill for R Markdown. The steps required to contribute a new article are as follows:

1. Fork the Distill template repository to host your article. We will use this repository for feedback and review before publishing your article.

2. Replace the article metadata fields in the template with information about your article.

3. If your article is based off of a report originally written in LaTeX, convert your TeX file to Markdown format and copy/paste text from the body of the converted Markdown file into your article template.

4. Make in-line edits as needed to meet our general formatting guidelines, and knit your .Rmd file. Commit and push the final version of the article to your Github repository.

5. Post an issue on the Data Science Student Projects GitHub repo to submit a request for publishing your article. Please include a link back to your article’s GitHub repo. Once we see your issue, we’ll take a look at your article, suggest changes as necessary, then publish it when it’s ready.

Article metadata

Your article should include some standard metadata fields. Here’s an example of a heading that includes all required metadata:

---
title: "Some Interesting Title Here"
description: |
  Using some fancy deep learning method to do something vaguely interesting 
  that is at least tangentially related to politics/computational social science. 
author:
  - name: Allison Koh 
  - name: Author 2
date: 07-17-2020
categories: 
  - Natural Language Processing
  - Machine Learning
bibliography: bibliography.bib
creative_commons: CC BY
repository_url: https://github.com/allisonkoh/project-repo
output: 
  distill::distill_article:
    self_contained: false
preview: figures/BERTfig3.png
---

Please note the following:

1. You can include multiple authors in the author field, and the field includes an option to include URLs if you would like to add a link to your personal website.

2. In the categories field, list the course(s) you took as a part of your project submission. Courses offered by the data science lab include: Machine Learning, Natural Language Processing, and Python Programming for Data Scientists.

3. The creative_commons field marks the article as being share-able (all contributed posts must have a Creative Commons license).

4. The repository_url is used to provide links from the article back to GitHub.

5. Select a preview figure that best represents your findings.

Converting TeX files to Markdown

If your article is based off of a report originally written in LaTeX, you can use pandoc to convert your article to the Markdown (*.md) format required We recommend using the {rmarkdown} package in R for file conversion.

Before file conversion, make sure you have the following documents in the same repository:

• LaTeX (*.tex) file
• Bibliography (*.bib) file
• Figures folder (NOTE: Make sure figures are saved as *.png files.)
• Any other cross-referenced materials

After compiling your article repo and ensuring you have the latest versions of {rmarkdown} and {knitr}, use the rmarkdown::pandoc_convert() function to convert your TeX file to Markdown (*.md) format. The code you run should look like this:

rmarkdown::pandoc_convert(
  "bertcnn-classifying-manifestos.tex", 
  to = "markdown", 
  output = "bertcnn-classifying-manifestos.md", 
  citeproc = TRUE
)

Once your *.md file has compiled, continue on from Step 3 in the instructions above.

Formatting Guidelines and Recommendations

You can use the following list of guidelines for your article to meet general formatting requirements. Please note that in-line edits necessary to publish the final version of an article will vary on a case-by-case basis.

• All section headings start with ##, and all sub-headings start with ###.
• Citations are formatted correctly¹.
• Figures are high-quality in resolution.
• Cross references to figures, tables, and sections are formatted correctly.
• (Most importantly) The .Rmd file knits with no issues.

In addition to these general guidelines, we also recommend converting tables into figures (full color) when possible.

Questions?

Please feel free to reach out at any point in the process of contributing your article for advice and feedback. You can direct questions about submitting your work to Allison Koh.

---

1. You might run into issues with bibliography files imported from some citation management programs (i.e. Mendeley).↩︎