Data Ethics Literature Review

This is research project to survey academic literature, courses, and other materials related to ethics in data science.

Very much a work-in-progress.

We use semantic web technologies to chart:

• Courses in data ethics
• Texts in data ethics
• Relations between the above

Goals

Goals of this research project include:

• production of a website to display a navigable collection of courses,
• production of a sample course syllabus

Structure of this Repository

• /aoir-abstract/: abstract for the project, for the AoIR conference.
• /data/: the main graph data, in Turtle-format RDF.
  • courseAndTexts.ttl: the main data file
  • courses.ttl: a subset of the above, missing bibliographic data
  • texts-hq.ttl: just a turtle version of the bibliography.bib file in the root directory
  • texts: graph data for texts, organized by course. Their format is {courseID}.{extension}.
    • txt: our manually-extracted references, copy-pasted from the syllabi
    • bib: bibtex files generated from those .txt files, using anystyle
    • ttl: turtle files generated from the bibtex files, using toRDF.py
• /notes/: an org-roam Zettelkasten, containing mostly hand-written notes for each source in the bibliography, and for additional concept notes, as needed.
• /papers: contains PDFs of papers themselves, if available. Each paper should be named according to its bibliographic key. These may be withheld from the Git repository, for the moment.
• /turtleize/: a set of Python scripts for generating the graph, manipulating the data, and visualizing it.
• bibliography.bib: a BibLaTeX file containing a manually-collected bibliography

Hacking

To run the code in this project:

1. Install the Nix Package Manager. See “Installing Nix,” below.
2. Enter the turtleize directory: cd turtleize
3. Run nix-shell, which will load the environment from default.nix.

See the README in the turtleize/ directory for more details.