Where To Next? A Dynamic Model of User Preferences

Passino, Francesco Sanna; Maystre, Lucas; Moor, Dmitrii; Anderson, Ashton; Lalmas, Mounia

doi:10.1145/3442381.3450028

Cited by 13 publications

(5 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This reduces the chance of including listeners who do not use Spotify as their primary means of listening to music. This does not create selection bias among Spotify's listeners in general (Sanna Passino et al 2021). As a check, we compared the track ages and genres streamed in this dataset to one in which users did not listen to at least one track every month.…”

Section: Methodsmentioning

confidence: 99%

Time after Time: Longitudinal Trends in Nostalgic Listening

Hanson

Anderton²,

Way³

et al. 2022

ICWSM

View full text Add to dashboard Cite

Nostalgia was once considered a medical disease but is now understood as a beneficial, identity-affirming emotion. Yet what induces feelings of nostalgia is not fully understood. Are nostalgic feelings prompted by changes in a person’s life, and do certain events prompt nostalgia across whole populations? In this paper, we analyze when people listen to nostalgic music. We use data from a large, cross-national survey to train a classifier to detect which tracks are nostalgic for individual listeners. We then analyze a comprehensive dataset of listening histories from a 5.5 year period. Despite it being a complex concept, we were able to predict nostalgic listening with relatively high precision. We compare our results across listeners from four countries to understand how consistent behavior toward nostalgic music is. We find people listen to nostalgic music more often as they age. We also find people tend to listen to nostalgic music consistently in their day-to-day lives. We do not find evidence that listening to personally nostalgic music increases in response to particular traditions, seasons, or events. However, we do find traditions and events can affect how much “back catalog” music people listen to. These trends are consistent across the national contexts we studied. Our results advance prior findings about nostalgia and the life course, and demonstrate a novel methodological consideration for studies of nostalgic listening.

show abstract

Section: Methodsmentioning

confidence: 99%

Time after Time: Longitudinal Trends in Nostalgic Listening

Hanson

Anderton²,

Way³

et al. 2022

ICWSM

View full text Add to dashboard Cite

show abstract

“…Studies on online music platforms reveal users' preferences evolve over time, with more diverse music consumption leading to higher engagement [1,30]. This diversity increase may represent online coherency maximizing exploration, contrasting with offline contexts where constant exploitation may lead to user boredom.…”

Section: User Decision Processmentioning

confidence: 99%

Uncovering the Deep Filter Bubble: Narrow Exposure in Short-Video Recommendation

Sukiennik,

Gao,

2024

Proceedings of the ACM Web Conference 2024

View full text Add to dashboard Cite

Filter bubbles have been studied extensively within the context of online content platforms due to their potential to cause undesirable outcomes such as user dissatisfaction or polarization. With the rise of short-video platforms, the filter bubble has been given extra attention because these platforms rely on an unprecedented use of the recommender system to provide relevant content. In our work, we investigate the deep filter bubble, which refers to the user being exposed to narrow content within their broad interests. We accomplish this using one-year interaction data from a top short-video platform in China, which includes hierarchical data with three levels of categories for each video. We formalize our definition of a "deep" filter bubble within this context, and then explore various correlations within the data: first understanding the evolution of the deep filter bubble over time, and later revealing some of the factors that give rise to this phenomenon, such as specific categories, user demographics, and feedback type. We observe that while the overall proportion of users in a filter bubble remains largely constant over time, the depth composition of their filter bubble changes. In addition, we find that some demographic groups that have a higher likelihood of seeing narrower content and implicit feedback signals can lead to less bubble formation. Finally, we propose some ways in which recommender systems can be designed to reduce the risk of a user getting caught in a bubble.

show abstract

“…Proactive recommendation is an emerging research domain that primarily comprises two main research lines: 1) Preference shifts in recommendation 2) user preference guiding . In the domain of preference shifts, prior works predominantly concentrate on understanding how user preferences evolve when interacting with recommender systems, often through simulation-based approaches [5,6,12]. Some studies also attempt to optimize long-term rewards rather than myopic behaviors under user preference shifts [3,7,9].…”

Section: Related Workmentioning

confidence: 99%

Proactive Recommendation with Iterative Preference Guidance

Bi,

Wang,

Pan

et al. 2024

Companion Proceedings of the ACM Web Conference 2024

View full text Add to dashboard Cite

Recommender systems mainly tailor personalized recommendations according to user interests learned from user feedback. However, such recommender systems passively cater to user interests and even reinforce existing interests in the feedback loop, leading to problems like filter bubbles and opinion polarization. To counteract this, proactive recommendation actively steers users towards developing new interests in a target item or topic by strategically modulating recommendation sequences. Existing work for proactive recommendation faces significant hurdles: 1) overlooking the user feedback in the guidance process; 2) lacking explicit modeling of the guiding objective; and 3) insufficient flexibility for integration into existing industrial recommender systems. To address these issues, we introduce an Iterative Preference Guidance (IPG) framework. IPG performs proactive recommendation in a flexible post-processing manner by ranking items according to their IPG scores that consider both interaction probability and guiding value. These scores are explicitly estimated with iteratively updated user representation that considers the most recent user interactions. Extensive experiments validate that IPG can effectively guide user interests toward target interests with a reasonable trade-off in recommender accuracy. The code is available at https://github.com/GabyUSTC/IPG-Rec. CCS CONCEPTS• Information systems → Recommender systems.

show abstract

Where To Next? A Dynamic Model of User Preferences

Cited by 13 publications

References 36 publications

Time after Time: Longitudinal Trends in Nostalgic Listening

Time after Time: Longitudinal Trends in Nostalgic Listening

Uncovering the Deep Filter Bubble: Narrow Exposure in Short-Video Recommendation

Proactive Recommendation with Iterative Preference Guidance

Contact Info

Product

Resources

About