Gender Diversity Analysis in OSS Projects

PyData Meetup, Madrid

Transcript

1. Gender Diversity Analysis in OSS Projects PyData Meetup 29th June, 2017 Madrid Daniel Izquierdo, CDO dizquierdo@bitergia.com @dizquierdo speakerdeck.com/bitergia
4. Tweet => 13% of people attending the OpenStack Summit were women Tweet => How many of them are actually contributing to the source code? Intro
5. Gender diversity of the technical contributions in the OpenStack Project Teams Problem
6. Goal-Question-Metric approach • Contextualize tech gender-diversity groups • Data sources available for the analysis • Tooling • Results and Further Work How To
7. Governance -> Goals <- Questions <- Metrics Goal: Increase gender diversity in the OpenStack Foundation Context
8. How’s performing the industry with this respect? Context
9. FOSS Survey in 2013: - 11% of women answered the survey The Industry Gender Gap by the World Economic Forum. - 5% for CEOs, 21% for Mid-level roles, 32% of Junior roles Context
10. Some companies https://blog.pinterest.com/en/our -plan-more-diverse-pinterest http://www.google.com/diversity/ http://newsroom.fb.com/news/2015/0 6/driving-diversity-at-facebook/ https://blogs.dropbox.com/dropbox/2014/11/stren gthening-dropbox-through-diversity/
11. Question: How’s performing OpenStack? We need data!!!! Context
12. Git: https://git.openstack.org Gerrit: https://review.openstack.org/ Others: Mailing Lists, Launchpad, IRC… >1M commits and > 1.5M code review votes Data
13. Git example: commit 61ab0c46b09299c07e86320d612a0fcc281491b1 Author: Daniel Izquierdo <dizquierdo@bitergia.com> Date: Fri Apr 28 00:04:05 2017 +0200 Add 'by default' values when eventizing Data
14. Tooling Original Data Sources Mining Tools Perceval @ GrimoireLab Info Enrich. Genderize.io Ceres/ Pandas Jupyter Notebooks Manual work Viz ElasticSearch + Kibana
15. Results Original Data Sources • Git and Gerrit repos based on yaml at Governance • ~ 1M commits • ~ 500K changesets • ~ 1.5M patchset uploads • ~ 1.8M patches code reviews
16. Results Mining Tools Perceval • At grimoirelab.github.io • Parses API’s, logs, etc and produces JSON documents • Those are later stored in ElasticSearch
17. Results Info Enrich. Genderize.io Pandas Jupyter Notebooks Manual work • Genderize.io: name database • Ceres: data analysis lib. to work with Perceval • Jupyter Notebook: web app. For data analysis • Manual work:
18. Results Viz ElasticSearch + Kibana • ElasticSearch: Schemaless db • Kibana: works great with ES • This tandem helps a lot to verify info • Drill down capabilities
19. Demo OpenStack Diversity Dashboard (private access)
20. Women activity (last year): ~ 11% of the population ( ~ 340 active developers ) ~ 9% of the activity ( >=6k commits ) OpenStack (Austin)
21. Women activity (last year): ~ 6.8% of the activity ( ~ 4k commits ) ~ 9.9% of the population ( ~ 330 active developers ) Linux Kernel
22. Women activity (last year): ~2K commits (6.5% of the activity) 71 developers (8.5% of the population) Hadoop
23. Users It’s important to understand your potential users! C-level? Middle management? Developers? Community? This study aims at understanding the current situation And look for best practices
24. Decisions based on data!
25. Bitergia Software Development Analytics for your peace of mind
26. Thanks! Daniel Izquierdo, CDO dizquierdo@bitergia.com @dizquierdo