Data Science of Messy Metrics

2,142 views
2,015 views

Published on

My #TEDxPoynter talk slides

Published in: Technology, Business
0 Comments
2 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
2,142
On SlideShare
0
From Embeds
0
Number of Embeds
81
Actions
Shares
0
Downloads
0
Comments
0
Likes
2
Embeds 0
No embeds

No notes for slide

Data Science of Messy Metrics

  1. 1. HelloGilad Lotan | @gilgul |gilad@betaworks.com
  2. 2. still sexy?
  3. 3. Illustration: Matt Taylor
  4. 4. status affordances
  5. 5. let’s look at some data
  6. 6. Establishing the Norm
  7. 7. Data Reflects Us
  8. 8. Social MovementTV ShowAwardsCeremony
  9. 9. Newtown ShootingNews BreaksObama visits
  10. 10. Volatility of Trends over Time
  11. 11. network analysis
  12. 12. User A User BGraph Representation of Social DataUser A is connected to BUser A User BUser A follows User BWord A Word BWord A appeared with Word BUser A User BUser A retweeted User Bconnectionfollowsappears withretweeted
  13. 13. Graph Measures• Centrality– Betweenness– Closeness– Eigenvector– Degree• Clustering Coefficient (clique)• Modularityoutdegree=2indegree=3degree=5
  14. 14. Twitter Users with “python” in theirBios• 2 days of Twitterdata• 4246 users• 62k tweets• Colors representmodularity class
  15. 15. Pythonistas onTwitterEnglish / EuropeanJapanesePython(the snake)ChineseSpanish SpeakersMusicians, Artists
  16. 16. #Debates / Ohio
  17. 17. #Debates / OhioPoliticosOSU StudentsOhio based Media
  18. 18. VP Debate
  19. 19. supportaccusations
  20. 20. Fragmented audiences+multiple platforms=very messy data
  21. 21. metrics in context
  22. 22. graph-based analytics
  23. 23. Gilad Lotan | @gilgulgilad@betaworks.com

×