2. WiNLP Coming to Life
ACL 2016 – Bonnie Webber
notices the underrepresentation
of women @ ACL conferences
EMNLP 2016 – 1st WiNLP
planning meeting – goal, format,
funding and organization
ACL 2016 – Margaret
Mitchell organizes the
WUGiNLP Lunch, sponsored
by Microsoft Research
ACL 2017 – 1st
WiNLP workshop
with over 130
participants!
3. Our Mission
• Help promote and support ideas and voices of women & other
underrepresented groups in Natural Language Processing
• Goals of WiNLP
▪ Increase awareness of the work of women & underrepresented groups
▪ Support women and underrepresented groups in continuing to pursue their
research
▪ Build a community of early-stage researchers of underrepresented groups
to support one another
▪ Motivate long-term resources for underrepresented groups within ACL
4. Our Mission - Deliverables
• Significantly increase the percentage of women and
underrepresented groups in the ACL community
• Create a directory of women and underrepresented
groups to increase awareness of who researchers are in NLP
• The foundation for a long-term diversity and inclusion
effort within ACL
5. Our Process - Worth Noting
• One-on-one writing mentorship
• Telepresence
• Career mentoring here today
• Recruiting & Sponsorship from top companies
6. Outreach/Initiatives at ACL
• Rotating locations
• Local conferences
▪ NAACL, EACL, AFNLP, Nodalida, etc.
• Funding opportunities
▪ Don and Betty Walker International Student Fund
• On-site childcare
▪ New at ACL 2017
7. ACL Report – Nomination Proc.
• Goal - examine ACL’s nomination and selection procedures
• Committee - Yejin Choi, MinYen Kan, Joakim Nivre (chair), and
Marilyn Walker
• Partial list of objectives
▪ ACL Fellows program
▪ Selection of general and program chairs for ACL conferences
▪ Lifetime achievement awards
9. ACL Executives Each Year
Report on ACL Nomination Procedures and ACL Fellows
Male
Female
10. ACL Report – Recommendation
• Nominating committee and LTA committee replaced by single
committee with diverse representation that is partly
independent of the executive committee
• But what do the numbers tell us about the participants and
authors at ACL events?
11. Diversity at the Main Event
• ACL 1979 – 28 papers, 20% female first authors
• ACL 2016 – 20% female first authors
Male
Female
Charts courtesy of Margaret Mitchell & Hal Daumé III
12. Diversity at the Main Event
Charts courtesy of Margaret Mitchell & Hal Daumé III
14. What can be Done?
• Following the example of successful initiatives
▪ Women in Data Science (WiDS)
▪ Women in Computer Vision (WiCV)
▪ Women in Machine Learning (WiML)
• 2006 – 100 participants, 60 posters
• 2016 – 570 participants, 231 posters
• Largest group on directory – NLP!!!
• Do these NLP researchers attend
ACL?
• What about other dimensions
of diversity?
15. How WiNLP can Help
• We are co-located with ACL, the biggest NLP conference
▪ Come see our poster session!
▪ Come listen to the talks!
• Mentoring and advice incorporated in the reviewing process
• Travel assistance for accepted authors
▪ 1st time to attend an international conference for 30% of the applicants
▪ 6% undergraduate students
16. Challenges and Barriers
Challenges of being outnumbered
• “In my workplace [Gong.io - an Israeli startup] I am the only
female engineer (among ~20 male engineers) and I hope
that within several years these statistics might change. If I can
take a part of this change, I'll be very proud.” – Lotem Peled
• “In computer science, it can be difficult being a women of
color where diverse representation is lacking.” – Candace Ross
17. Challenges and Barriers
Bias against underrepresented minorities in NLP
• “Surprisingly, my education is problematic for many, some inferring
that I kept switching degrees because I was not very good at any
one thing or was confused. I’ve even got “her background is, well…
broad” as a dismissing and the only remark in a formal presentation
of my research in front of group of peers. I believe that the
perception of my academic achievements has been negative
because I am a woman, black or both. Rather than being a
couple random examples of our careers as minorities, such
incidents are (always surprisingly) frequent.” - Natalie
Schluter
18. Challenges and Barriers
Challenges of being a woman in NLP
• “One of the strategies I have seen female computer scientists use to
try and fit in is, intentionally or unintentionally, hiding
preferences or characteristics thought of as a being
stereotypically female to feel "included" in a male-centric
community. For instance, following the stereotype of a hardworking
male researcher, staying in the lab until late despite being tired and
pretending not to be interested in anything apart from work. On a
superficial level this works, but might lead to a feeling of losing
one’s true self.” – Yuki Arase
19. Challenges and Barriers
Challenges of not working in a big US lab
• “Groups that are based in North America have more media
exposure and access to policy makers. We have to make more
of an effort to affect impact on the global stage. It can be done,
however, through persistence.” – Pascale Fung
• “One of the challenges I experience is the difficulty in
establishing collaborations. Unfortunately, in Algeria, only a few
people work on NLP. It serves as real barrier trying to find someone
to share our ideas or simply communicate about a given problem.” –
Imane Guellil
20. Challenges and Barriers
Challenges of not working in a big US lab
• “Lack of money for conducting research and connecting with
the overseas research community are also limited for us [in
Vietnam]. As a result, I just attended international conferences
that were held in my country because the cost of travel to
conferences, like ACL+WiNLP, is nearly my salary for one
year as a lecturer.” – Ngo Thi Lan
21. Challenges and Barriers
Bias against certain research areas
• “One of the biggest challenges NLP professionals currently face, in my
experience, is that algorithms and mathematical approaches,
basically quantitative approaches, receive most of the credit in
NLP research. I see this bias in favor of mathematical solutions as a type
of barrier in the NLP community.” – Martha Palmer
• “One big challenge I see in NLP is that there seems to be an almost
exclusive focus on task performance, measuring the quality of the
models using abstract metrics. Analysis studies are often undervalued and
difficult to publish, although thorough, insightful analyses are valuable and
difficult to carry out.” – Dong Nguyen
22. Challenges and Barriers
General challenges of academia
• “The biggest challenge in NLP (and academia in general) is
how to maintain high-quality peer review and reward
quality work beyond numbers of publications, especially
when the field is exploding.” – Wei Xu
23. How WiNLP can Help
• But, certain challenges apply not only to women and
underrepresented minorities…
▪ True, yet even more important here
• Does WiNLP address these challenges?
▪ Yes! Mentoring, funding, telepresence, co-location with main event
24. WiNLP Submission Process
• 94 submitted abstracts
▪ Abstracts are not archived
▪ Enables submission of previously published work, work-in-progress
▪ Increase visibility in the conference
• 75 accepted abstracts
• 16 papers rejected
• 4 abstracts withdrawn
Thank you to our amazing
reviewing committee!!!
28. WiNLP Program
• 3 invited talks
• 8 contributed talks
• Over 60 posters
• 2 mentoring sessions
▪ WiNLP registered participants, sponsors, and mentors
29. Mentoring Sessions
• Lunch round-table event
▪ Salon 3 – right next door
▪ Grab a plate and find a table – topic-based and sponsorship tables
▪ Move around
• 1-on-1 mentoring
▪ Introduction emails were sent to mentors/mentees
▪ Matching based on research interest
▪ Meet here after the talks or during the day
• Not a mentor yet? Volunteer for next year’s event!
35. What’s Next?
•Will there be WiNLP 2018? Where will it be?
• How can you be a part of WiNLP/ become a mentor?
•What else is planned?
Come to the concluding remark to hear all about it!