Wikiversity:Fellow-Programm Freies Wissen/Einreichungen/Platform Governance Archives/Projekttagebuch

Fellow-Programm Freies Wissen

Start

Offene Wissenschaft

Programm

Fellows

Mentoring

Advisory Board

Veranstaltungen

Blog

Projekttagebuch (Platform Governance Archives) Project's main milestones (as of November 2020) Conclude the data collection of all available versions of all policies of the ten social media platforms mentioned in the project, using both the Wayback Machine and platforms' websites. Expected completion date: January 2021. Translate the PDF files into Markdown files, effectively creating two mirror databases. Expected completion date: January 2021. Conclude the manual tagging of all documents (or part of them), classifying them by topic (e.g. hate speech, copyright, spamming, privacy etc). Expected date: March 2021. Define what exactly the data visualisation tool will do. Expected date: March 2021. Define the language and tools that will structure the website, as well as how to host it. Expected date: March 2021. Implement a beta version of the website. Expected date: April 2021. Test the usability of website, tweaking it if needed. Expected date: April 2021. Define a communication plan for the website. Expected date: May 2021. Launch the website. Expected date: unknown. November 2020 Based on our October assessment, we decided to re-consider our data collection methods and are currently experimenting with Wayback Machine's API, with the help of another HIIG colleague, Tom Sühr. Since this is largely a new territory for us, we are going to use YouTube as a case study to test our methods. Meanwhile, the manual reconstruction of some highly fragmented policies (e.g. Community Guidelines) is underway. October 2020 In October, most of my work concerned reviewing and identifying issues with the collection of documents that had already been assembled for the project. Together with my HIIG colleagues Adrian Kopps and Christian Katzenbach, I realised that much of what counts as "platform policies" was either not yet covered by our database or archived in such a fragmented way that no analysis could be done. This was an important realisation, and led us to reconsider which techniques (and skills) were needed to develop the PGA.