About the Author :: About this Website :: Legal and Credits :: Pelican :: Privacy
About the Author
My name is Jacob Levernier. I am passionate about coming to some sensible answers about doing data science in an ethical way. I am also the CLIR Bollinger Postdoctoral Fellow in Library Innovation at the University of Pennsylvania, My position at the University of Pennsylvania is akin to a Concept Designer's in the arts, but for library technology and data mining techniques. an amateur mnemonist, and a student of the Liberal Arts.
My field of interest is "Data Science," at the intersection of Psychology, Data Analysis and Statistics, Computer Science, and Library and Information Science. My research interests include moral development and data literacy — How do people best understand the implications of (and make decisions based on) data ethics issues, and what influences their beliefs about those issues? In line with these interests, I'm involved with efforts to write data analysis policy that explicitly considers questions of ethics and enables users to be a part of the conversation with analysts and administrators around the access and use of the data they generate.
Data science is exciting. Seeing everything as data is fundamentally paradigm-shifting; each new project and devised analysis approach feels like an object placed in an antique Wunderkammer. Each excites the imagination and prompts new inquiry. In my experience, social scientists typically have very little experience with any sort of programming, and programmers typically have very little experience with statistics or training in research ethics. Neither group tends much to connect the history of ethics to ongoing investigations. History, I think, can ably found future inquiry. Old words can ring true in new and formative ways. The ideas of those who have come before us, especially outside of our own fields of specialty, can push us forward to develop not only sound methodologies for pursuing the answers we seek but also innovative questions through which to begin that pursuit. This field is new, its questions are increasing in their relevancy, and it lends itself to an interdisciplinary approach.
About this Website
The title of this site, Ad Unum Datum, means, in loose Latin, "down to a single piece of data."
This website exists to serve three purposes. First, it is meant to be an archive of educational and research activities on data ethics and practices. Second, it is meant to be a Commonplace Book for things that ignite my curiosity. Third, it is meant to stimulate collaborative dialogue among scientists, programmers, engineers, artists, and others.
All material that has been quoted or otherwise posted on this website and is not my original work is indicated as such and is intended to aid in educational projects. If you are the original owner of any posted work and believe that it's not being used in a way that complies with Fair Use principles, please contact me.
I keep an up-to-date list of links and credits, especially for images and icons, here.
The photograph at the top of this page is courtesy of Breathe Photography in Eugene, OR, and was taken on Halloween in 2013 at a fradulent seance performance. It's a topic in which I have a hobbyist's interest.
Short bits of code that do not have a license attached to them (e.g., snippets of CSS) are for you to use without feeling like you need to give credit (although it's always nice): See the CC0 license. More substantive bits of code may have alternative (non-CC0) license notes attached to them, usually in their headers.
Any text not otherwise cited is © Jacob Levernier and released under a Creative Commons BY-NC-SA license (if you would like to use it in some other way, though, feel free to contact me to arrange a different license).
I've used this much text to describe license terms because I am a proponent of best practices for data and code management, and because I worry that leaving legally-relevant expectations ambiguous can lead to one or both parties in an interaction (here, the author and the user) feeling cheated. When someone authors or otherwise creates code, icons, or other types of content, and says it's "free," it's often unclear what "free" means (Does "free" mean that I can use it, but need to provide a link back to the author's website? Does it mean that I can use it as long as I'm not profiting off of its use?). It's even more problematic when an author states that something is "free to use," but has an "All Rights Reserved" notice at the bottom of the page (those two statements are potentially at odds with one another). Thus, in contrast to some who purposely leave their work unlicensed, I prefer to make all copyright and license terms as explicit as possible.
The website is built with Pelican and the Python SCSS package. The theme is built on Matheiu Agopian's Mnmlist theme, which Mr. Agopian generously released to the public waiving copyright. Portions of the theme are also built on the Pelican dev-random2 theme, which is released under similarly permissive terms. In addition, the site benefits from the Tipue Search jQuery plugin, as well as Talha Mansoor's AGPL-licensed Pelican plugin of the same name. It also benefits from the Math Render Plugin for Pelican, which enables the use of MathJax for rendering mathematical formulae with \(\LaTeX\).
This website uses a privacy-enhancing Pelican plugin that I wrote. It also uses a plugin that I created for better line numbering in code blocks.
Motoki Wu proposed the idea of a textual gloss here. Having somewhat of a background in Classics and mnemotechny (in the tradition of which mnemonic glosses were prevalent), this idea appeals to me. I'm grateful to use it. I've added details on my implementation of this idea here. It allows using shortcodes and adding images, like this one (click it and hold to enlarge):
I use an open-source, self-hosted analytics platform called Piwik to track statistics for this website. This analytics engine employs several features to increase user privacy while still giving me useful information about site usage:
- Do Not Track browser preferences are honored.
- URLs can be made pretty through a self-hosted site, links.AdUnumDatum.org, rather than a third-party site such as bitly. These redirects are powered by a single PHP script, which also records limited analytics using Piwik.
- IP addresses are semi-anonymized: Rather than recording a visitor's full IP address (e.g., '192.168.0.1'), only the first two bytes are recorded (e.g., '192.168.*.*').
I've listed these points not because I think it likely that most users will care, but precisely because I think that most users won't care. As with the the YouTube plugin mentioned above, I think that making privacy-relevant information explicit is good practice and worth doing if for no other reason than to increase awareness of tracking issues.
If you feel like the protections above aren't sufficient and would like to opt-out of tracking, you can uncheck the box below. Please note that the opt-out box will display only in browsers that support iframes. Some mobile browsers don't. Normally I would advocate keeping choices like this "opted-out" by default. However, I've tried here to balance a concern for privacy with a need for site analytics.