AnaAndreea Stoica Christopher Riederer Augustin Chaintreau Columbia University The effects of social recommendations on network diversity Highlights of Algorithms June 2018 Automating social decisions ID: 776440
Download Presentation The PPT/PDF document " Algorithmic Glass Ceiling in Social Net..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Algorithmic Glass Ceiling in Social Networks
Ana-Andreea Stoica, Christopher
Riederer, Augustin ChaintreauColumbia University
The effects of social recommendations
on network diversity
Highlights of Algorithms, June 2018
Slide2Automating social decisions
2
Slide3Automating social decisions
3
Slide4Automating social decisions
4
Slide5Mirroring social biases
5
Slide6Mirroring social biases
6
Slide7Mirroring social biases
7
Slide8Automating social decisions
8
Recommendation systems
Slide9Automating social decisions
9
Recommendation systems
Slide10“The glass ceiling effect is the unseen, yet unbreakable barrier that keeps minorities and women from rising to the upper rungs of the corporate ladder, regardless of their qualifications or achievements.”
Federal Glass Ceiling Commission. "Solid investments: Making full use of the nation's human capital. US Government, Department of Labor. Washington, DC: US Government Printing Office. Retrieved February 2013." (1995).
10
Slide11Glass ceiling effect
11
Data collected by LeanIn.Org and McKinsey & Co. from 134 global companies, surveying 34,000 men and women. http://graphics.wsj.com/how-men-and-women-see-the-workplace-differently/Shirin Nilizadeh et al. "Twitter's Glass Ceiling: The Effect of Perceived Gender on Online Visibility. AAAI Conference on Web and Social Media. 2016.Avin, Chen et al. "Homophily and the glass ceiling effect in social networks.” ITCS, ACM, 2015.
Only a handful of
women make it to the
C-Suite.
Males have
more retweets and followers.
% females decreases for more prominent researchers.
Slide12Empirical results on Instagram Data
Graph of likes and comments from Instagram public profiles split by gender
Simulated two recommendation algorithms:Adamic-AdarRandom walk of length twoGlass ceiling effect is amplified
12
Slide13Model
Ingredients for a glass ceiling effect: Minority-majority: blue label and red labelFraction of red nodes = r < ½Rich-get-richer: nodes connect w.p. proportional to degreeHomophily: if different labels, connection is accepted w.p. ρ
13
Organic growth
Recommendation model
Slide14Degree distribution
Recommendation model:
Organic growth:
Theorem
: For 0 < r < ½ and 0 ≤
ρ
≤ 1, for the graph sequences G(n) for the organic model and G’(n) for the recommendation model, the red and
blue
populations exhibit a power law degree distribution with coefficients:
14
gap
Slide15Key takeaways
Seemingly neutral algorithms may reinforce inequality Trade-off between prediction rate and fairnessAlgorithm design with awareness of network structure
15
Slide16Thank you!
Web: http://www.columbia.edu/~as5001/
Email: as5001@columbia.edu