Data Science How do you Even?
summary
Claims concrete wins: an implemented quality filter yields a 40% reduction in blocks following @mentions from accounts users don’t follow, a timeout restriction produces 25% fewer abuse reports for offenders, and suspension evasion detection leads to three times more abusive accounts suspended within 24 hours and >50% of enforced abusive content proactively surfaced. Teams prototype metrics with ad hoc pipelines, build ML prioritizers for reports, and productionize measures with data engineers to measure impact (16% fewer abuse reports after certain interactions).
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:15Speaker's background
- 6:00What is Data Science?
- 8:35Data Science @ Twitter
- 11:02Health Data Science
- 15:43Potentially abusive content and User Controls
- 22:48Suspension evasion
- 26:15Metric Development and Production-ization
related talks
52mAB-tests, lies, damned lies, and statistics.
31mThe Hidden Costs of Data Quality - Tackling Common Data Challenges in ML
51mFairytale vs. Narrative: Telling stories and making decisions with data
08mdata science in automotive
43mKeynote
21mData Analytics for Better Product Decision Making
52mMaking Better Decisions with Data
08mAI-Driven Innovation for Enhanced Food Delivery Operations