Data, Privacy, & Trust in Open Source: 10 Lessons from Wikipedia
summary
Wikimedia Foundation recasts privacy around concrete practices: builds an event logging JSON schema published on the Wiki so every tracked field is explicit, runs client/server A/B tests but offers an opt-out for A/B tests and beta features, and purges sensitive logs (IP checkuser data) on a 30-day IP purge schedule. A 2006 search-data release scare and a 500-person public policy review (vs ~75,000 active editors) drive the rule: collect only what answers a question, aggregate quickly, and delete.
AI-written summary · about
chapters (24)
- 0:00Intro
- 0:57Data Privacy Trust
- 3:15Editor Retention
- 4:18Oh Shit Graph
- 4:53User Stories
- 5:15Lean Methodology
- 7:47Balancing User Stories
- 8:41Privacy Policies
- 9:36TOS Dr
- 10:28Privacy Checkup
- 12:01Privacy Standards
- 13:03Ownership
- 13:51Data Pipeline
- 15:12Data as a Tool
- 21:34Purging Aggregating Data
- 24:08No Real Anonymization
- 26:35DeepWiki Inspector
- 28:42Why Wikipedia
- 33:34Collecting Schema Data
- 34:39Percentage of Likes
- 36:04Privacy
- 37:44Multilingualism
- 39:25Privacy Expectations
- 40:35Editing via Tor
related talks
46mPrivacy, data protection, and open source development
10mLightning Talk: How Not to Make Open Source - Lessons Learned
08mOpen Source Developers Make Tech More Trustworthy
31mUsing the Commons without causing a tragedy
27mA Guide to Responsible Data Collection In Open Source
41mWikiLoop: Big Tech's Open Knowledge contributions
11mOpen Source Science - Science and Sharing
35mRegulations, Ethics And Messy Data: Tales From An Open Source Analyst