Exploring Author Gender in Book Rating and Recommendation Michael - - PowerPoint PPT Presentation

▶
exploring author gender in book rating and recommendation
SMART_READER_LITE
LIVE PREVIEW

Exploring Author Gender in Book Rating and Recommendation Michael - - PowerPoint PPT Presentation

Sep 04, 2023 274 likes •455 views

Exploring Author Gender in Book Rating and Recommendation Michael D. Ekstrand People and Information Research Team, Boise State University Mucun Tian People and Information Research Team, Boise State University Mohammad R. Imran Khazi Texas

slide-1
SLIDE 1

Exploring Author Gender in Book Rating and Recommendation

Michael D. Ekstrand

People and Information Research Team, Boise State University

Mucun Tian

People and Information Research Team, Boise State University

Mohammad R. Imran Khazi

Texas State University (alum)

Hoda Mehrpouyan

Boise State University

Daniel Kluver

MacalasterCollege

slide-2
SLIDE 2

Diversity and Representation in Book Authorship

2

slide-3
SLIDE 3

3

Source: Canadian Women in the Literary Arts. http://cwila.com/2015-cwila-count-methods-results/

slide-4
SLIDE 4

4

How do recommender systems interact with these efforts?

slide-5
SLIDE 5

Hurdles by Ragnar Singsaas, used under CC-BY-SA 2.0. https://flic.kr/p/5jgjJP 5

slide-6
SLIDE 6

Research Questions

rq1 How are author genders distributed in cataloged books? rq2 How are author genders distributed in user book ratings? rq3 How are author genders distributed in recommendations? rq4 How do recommendations respond to user profiles?

6

slide-7
SLIDE 7

Fairness Positioning

Provider fairness (sort-of…) [Burke 2017] Calibrated fairness [Steck 2018] Descriptive, not normative

7

slide-8
SLIDE 8

Data

8

Ratings BookCrossing Amazon Books OpenLibrary LoC Authors VIAF ISBN Name

slide-9
SLIDE 9

rq1: Catalog Distribution

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% LoC Amazon BX (Explicit) BX (All)

Book Gender

Female Male Ambiguous Unknown Unlinked 9 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% LoC Amazon BX (Explicit) BX (All)

Book Gender (Known Gender)

Female Male

slide-10
SLIDE 10

Recommender Experiment

1. Sample 1000 users (each rating 5 books with known author gender)

  • 2. Measure user profile gender distribution (rq2)

3. Generate 50 recommendations for each user

1. User-User 2. Item-Item 3. MF (Funk SVD) [didn’t personalize - ignore] 4. Poisson factorization

  • 4. Compute recommendation list distribution (rq3)
  • 5. Compare recommendation lists to user profiles (rq4)

10

slide-11
SLIDE 11

Hierarchical Bayesian Model

11

𝑜𝑣 𝑧𝑣 𝜄𝑣 ෤ 𝑧𝑣𝑏 𝜗𝑣𝑏 ෨ 𝜄𝑣𝑏 ෤ 𝑜𝑣𝑏

Regression (in log odds) Rec List Balance (% Female) User Profile Balance (% Female)

Data Inferred

slide-12
SLIDE 12

rq2: Profile Distribution

Mild tendency towards male authors (mean < 0.5) High variance in user profile composition Average is more balanced than book catalog

12

slide-13
SLIDE 13

rq3: Recommendation List Distribution

13

Less variance than user profiles Average balance usually comparable Nearest-neighbor had most shift (U-U on explicit ratings, I-I on BX)

slide-14
SLIDE 14

rq4: Recommendation List Response

14

Input balance propagates, though extent varies

slide-15
SLIDE 15

Limitations

  • Rating data is extremely sparse
  • Algorithms didn’t perform particularly well
  • MF very non-personalized
  • Only considers binary gender identities
  • Working on statistical models to overcome that
  • Just a few algorithms

Philosophy: expand knowledge with what we have, work on the limitations.

15

slide-16
SLIDE 16

Conclusion

Summary

  • Users exhibit mild, diffuse

tendency towards male authors

  • User profiles more balanced than

book catalog

  • Nearest-neighbor & PF algorithms

propagated (some) user balance to recommendations

FutureWork

  • Better data
  • Better statistical model
  • More author features
  • More domains
  • More algorithms
  • Study diversifying algorithms

16

Questions?