Data Science for Workshop Organisation

Selecting participants for workshops or speakers for conferences is hard. Usually, there are more qualified applicants than there are free spots. At the same time, organisers are often interested in diversifying their participants or speakers: in terms of topic, in terms of seniority, but also perhaps in terms of gender and ethnicty. This short talk presents a new open-source tool that helps making participant selection for conferences and workshops more transparent and can help reducing biases in the selection procedure.

Transcript

1. ! Tiana_Athriel " dhuppenkothen Data Science for Workshop Organisation — Automating Participant+ Talk Selection — Daniela Huppenkothen NYU Center for Data Science
2. #AstroHackWeek 2015* *www.astrohackweek.org
3. 168 applicants 50 free spots
4. diversity* important
5. diversity* important *academic seniority + subfield, coding + statistics skills, gender, ethnicity, …
6. How do we select?
7. #DataScience!
8. discrete optimisation problem! Thanks to: Brian McFee
9. Entrofy*
10. Entrofy* *https://github.com/dhuppenkothen/entrofy
11. 1) Pick categories of interest
12. 2) Collect Data
13. 3) Pick acceptable candidates
14. 4) Define targets within categories
15. 5) Pick number of participants to be selected
16. 6) Run code!
17. 7) Invite participants
18. Advantages: transparency and accountability
19. Entrofy*
20. Entrofy* *https://github.com/dhuppenkothen/entrofy