Exploring the filter bubble: the effect of using recommender systems on content diversity

By Tien T. Nguyen, Pik-Mai Hui, Max Harper, Loreen Terveen, J. Konstan. 2014

634 citations recorded by the source corpus.

Subfields: Information Systems, Management Information Systems, Sociology and Political Science, Recommender Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Systems

1. Bamshad Mobasher
2. Robin D. Burke
3. Himan Abdollahpouri
4. Lucia D'Acunto
5. Pushmeet Kohli
6. David Krueger
7. Emilee J. Rader
8. András György

In Management Information Systems

1. Silvia Chiappa

In Sociology and Political Science

1. Ray Jiang
2. Pushmeet Kohli
3. Emilee J. Rader
4. Rebecca Gray
5. Bamshad Mobasher
6. Robin D. Burke
7. Natali Helberger
8. Kate Starbird

In Recommender Systems

1. Bamshad Mobasher
2. Robin D. Burke
3. Himan Abdollahpouri
4. Silvia Chiappa

Citations in the corpus

• Degenerate Feedback Loops in Recommender Systems (2019, methodology): “Through an analysis on the MovieLens dataset, Nguyen et al. (2014) found that the diversity of items recommended, and those users engage with, gets narrower over time.”
• Hidden Incentives for Auto-Induced Distributional Shift (2020, background): “Nguyen et al. (2014) similarly study the effect of recommender systems on individual users’ content diversity, and find that the set of options recommended narrows over time.”
• Explainable artificial intelligence: A survey (2018, neutral): “Content recommendation [4] and generation of fake content [5] coupled with other technologies will deeply impact the social dynamics [6].”
• Exposure diversity as a design principle for recommender systems (2018, background): “…can, at least in principle, also be designed to complement a personal media diet, or expose viewers to opposing viewpoints, and thus not only give users the information that they have been looking for, consider relevant to them or are most likely to read (Nguyen et al., 2014; Stark, 2009).”
• Controlling Popularity Bias in Learning-to-Rank Recommendation (2017, background): “Systems that use active learning to explore each user’s pro€le will typically need to present more long tail items because these are the ones that the user is less likely to have rated, and where user’s preferences are more likely to be diverse [12, 15].”
• Examining Trolls and Polarization with a Retweet Network (2018, background): “Other research has examined filter bubbles at length, including in the contexts of diverse platforms—such as web searches [9] and recommendation systems [12]—and conversations, for example climate change [3], the 2016 Brexit referendum [5], gun control and the Sandy Hook Elementary School shooting [”
• Understanding User Beliefs About Algorithmic Curation in the Facebook News Feed (2015, methodology): “[25] analyzed a MovieLens dataset and found that both the diversity of items recommended by the system and items rated by users did indeed become less diverse over time, but only slightly.”