Getting to the Core of Algorithmic News Curators: A Case Study of - - PowerPoint PPT Presentation

▶
getting to the core of algorithmic news curators a case
SMART_READER_LITE
LIVE PREVIEW

Getting to the Core of Algorithmic News Curators: A Case Study of - - PowerPoint PPT Presentation

Nov 24, 2022 176 likes •323 views

Getting to the Core of Algorithmic News Curators: A Case Study of Apple News Jack Bandy (Northwestern University) @jackbandy 2 My Research Ideation Machine (Beta) (platform or technology) (vice or trait) Facebook Google Search antisocial

slide-1
SLIDE 1

Getting to the Core of Algorithmic News Curators: A Case Study of Apple News

Jack Bandy (Northwestern University) @jackbandy

slide-2
SLIDE 2

2

slide-3
SLIDE 3

My Research Ideation Machine (Beta)

Facebook Google Search Netflix Google Ads GPS YouTube Reddit Apple News

Is making us

antisocial intolerant biased impatient naive uninformed

?

(platform or technology) (vice or trait)

3

slide-4
SLIDE 4

Apple News

4

slide-5
SLIDE 5

Related Work

  • Apple News
  • Columbia Journalism Review

(Brown)

  • New York Times (Nicas)
  • Algorithm Audits (Sandvig)
  • Underrepresentation (Kay)
  • Filter Bubbles (Bakshy)
  • News Platform Audits
  • Google News (Haim;

Nechushtai)

5

slide-6
SLIDE 6

Research Questions

  • How do news curation systems like Apple News

influence the public’s media intake?

  • What is the system’s mechanism?
  • How often does it update?
  • Does it localize or personalize?
  • What content does it direct attention to?
  • What sources does it feature?
  • What topics does it feature?

6

slide-7
SLIDE 7

Black Box Dilemma

  • Proprietary code
  • No public APIs or endpoints
  • SSL Pinning
  • Possible data collection

methods:

  • Apple News Twitter (Brown)
  • Email Newsletters (Brown)
  • Crowdsource

7

$ ./scrape_apple_news ERROR $ ./scrape_apple_news ERROR $ ./scrape_apple_news ERROR

slide-8
SLIDE 8

Methods: The Crowd

  • Amazon Mechanical Turk
  • Pros
  • Circumvents black box
  • Real-world data
  • High parallelism/throughput
  • Cons
  • Data Verification
  • Inconsistent coverage

8

slide-9
SLIDE 9

Methods: Appium

  • Automated App Control
  • Pros
  • Lower cost
  • Sustained coverage
  • No manual inspection
  • Cons
  • Single channel
  • Data points in vitro

9

slide-10
SLIDE 10

Findings: Source Concentration

10

Relative Distribution of Trending Stories Combined: December 12th-20th, 2018; January 4th-12th, 2019

Fox News CNN People HuffPost Politico Newsweek BuzzFeed Vanity Fair Vox Washington Post

% of Trending Stories (n=576) 5 10 15 20 25

slide-11
SLIDE 11

Findings: The Human Touch

11

Trending Stories Top Stories Curation Algorithmic Editorial Staff Headlines Displayed 4 (6 on big screens) 5 Localization National National Personalization No No

  • Avg. Stories / Day

31 16 Total Stories 279 144 Total Sources 28 40

  • Avg. Stories / Source

9.9 3.6

  • Stdev. Stories / Source

14.6 3.3 #1 Source % 20.1% (Fox) 9.0% (WaPo) #1-#3 Sources % 50.5% 25.7% #1-#10 Sources % 85.7% 55.6%

Data collected January 4th-12th, 2019

slide-12
SLIDE 12

Conclusions & Next Steps

  • How does Apple News affect Local and Regional news outlets?
  • How do people actually use the app? Do they prefer one section?
  • Do similar patterns (source concentration, the human touch) show

up in other aggregators?

  • Have ideas? Reach out! @jackbandy jackbandy.com

12

Computational Journalism Lab

slide-13
SLIDE 13
  • Sandvig, C., Hamilton, K., Karahalios, K., & Langbort, C. (2014). Auditing Algorithms: Research Methods for

Detecting Discrimination on Internet Platforms. In Data and discrimination: converting critical concerns into productive inquiry (pp. 1--23). Retrieved from https://pdfs.semanticscholar.org/ b722/7cbd34766655dea10d0437ab10df3a127396.pdf

  • Kay, M., Matuszek, C., & Munson, S. A. (2015). Unequal Representation and Gender Stereotypes in Image

Search Results for Occupations. In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems - CHI ’15 (pp. 3819–3828). https://doi.org/10.1145/2702123.2702520

  • Bakshy, E., Messing, S., & Adamic, L. A. (2015). Exposure to ideologically diverse news and opinion on

Facebook (supplementary materials). Science, 348(6239), 1130–1132. https://doi.org/10.1126/science.aaa1160

  • Haim, M., Graefe, A., & Brosius, H. B. (2018). Burst of the Filter Bubble?: Effects of personalization on the

diversity of Google News. Digital Journalism, 6(3), 330–343. https://doi.org/10.1080/21670811.2017.1338145

  • Nechushtai, E., & Lewis, S. C. (2019). What kind of news gatekeepers do we want machines to be? Filter

bubbles, fragmentation, and the normative dimensions of algorithmic recommendations. Computers in Human Behavior, 90, 298–307. https://doi.org/10.1016/j.chb.2018.07.043

  • Brown, P

. (2018). Study: Apple News’s human editors prefer a few major newsrooms. Columbia Journalism

  • Review. Retrieved from https://www.cjr.org/tow_center/study-apple-newss-human-editors-prefer-a-few-major-

newsrooms.php

  • Nicas, J. (2018). Apple News’s Radical Approach: Humans Over Machines. New York Times. Retrieved from

https://www.nytimes.com/2018/10/25/technology/apple-news-humans-algorithms.html 13