[Air-L] Framework for ethical and reproducible social network experiments

Luke Hutton lh49 at st-andrews.ac.uk
Fri Sep 18 06:47:24 PDT 2015


Hi all,

I wanted to let you know about some tools we've recently released aimed at researchers conducting studies which use data from social network sites, which is hopefully of interest to some on this list!

PRISONER <http://prisoner.cs.st-andrews.ac.uk> aims to simplify the development of such studies, while making them more ethical and reproducible, with a few main benefits:

* PRISONER provides a consistent API for reading and writing social network data, and handles implementation details such as authentication so you can focus on building experiment-specific logic, and supports re-use of code which targets one service to be used against another.

* By expressing the data-handling requirements of your study in a declarative format, PRISONER is able to generate consent forms and supporting documentation for ethics board/IRB approval, as well as enforcing these constraints at runtime to ensure only the data needed for an experiment are collected, while also providing automated sanitisations to reduce the sensitivity of collected data, as necessary.

* PRISONER supports reproducibility of experiments by providing easily transferable artefacts to allow others to reproduce the methodology and ethical conduct of your study. In addition, automated mappings between deprecated versions of APIs allow code which targets older versions of social network APIs to be usable in the future, making social network research more sustainable.

An early release of PRISONER is now available to allow social media researchers to experiment with and evaluate the tools, and to help us steer the development of the framework to meet the needs of the community. We are keen to collaborate on projects which are interested in adopting PRISONER, to provide support and any necessary extensions to the framework.

We provide the source code, documentation, and Docker images of PRISONER and a working PRISONER example to help get up to speed. You can learn more from http://prisoner.cs.st-andrews.ac.uk and get updates from @EthicsPRISONER on Twitter. Feel free to contact me directly with any questions or comments.

Cheers,
Luke

---
Luke Hutton
Research Fellow, School of Computer Science
University of St Andrews

Twitter: @barneyboo


More information about the Air-L mailing list