Contextual and Sequential User Embeddings for Large-Scale Music Recommendation

Hansen, Casper Worm; Hansen, Christian; Maystre, Lucas; Mehrotra, Rishabh; Brost, Brian M.; Tomasi, Federico; Lalmas, Mounia

doi:10.1145/3383313.3412248

Cited by 86 publications

(39 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Session-based models have aimed at learning users' intent within a session [44]. They focus on capturing instantaneous users' interests to provide them with the best timely item match [16,33]. Sequential recommender systems exploit the sequential nature of user-item interactions [12,33,44,45].…”

Section: Related Workmentioning

confidence: 99%

“…As seen here, the majority of the listening events happen during the afternoon and evening hours. As previous works have noted [16,26,42], the time of the day could indicate users' interests, tastes, and budget in interacting with an app or listening to music. Therefore, it has a potentially significant impact on users' streaming content.…”

Section: Exploratory Analysismentioning

confidence: 99%

See 1 more Smart Citation

Variational User Modeling with Slow and Fast Features

Fazelnia

Simon

Anderson

et al. 2022

Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining

Self Cite

View full text Add to dashboard Cite

Recommender systems play a key role in helping users find their favorite music to play among an often extremely large catalog of items on online streaming services. To correctly identify users' interests, recommendation algorithms rely on past user behavior and feedback to aim at learning users' preferences through the logged interactions. User modeling is a fundamental part of this large-scale system as it enables the model to learn an optimal representation for each user. For instance, in music recommendation, the focus of this paper, users' interests at any time is shaped by their general preferences for music as well as their recent or momentary interests in a particular type of music. In this paper, we present a novel approach for learning user representation based on general and slow-changing user interests as well as fast-moving current preferences. We propose a variational autoencoder-based model that takes fast and slow-moving features and learns an optimal user representation. Our model, which we call FS-VAE, consists of sequential and non-sequential encoders to capture patterns in user-item interactions and learn users' representations. We evaluate FS-VAE on a real-world music streaming dataset. Our experimental results show a clear improvement in learning optimal representations compared to state-of-the-art baselines on the next item recommendation task. We also demonstrate how each of the model components, slow input feature, and fast ones play a role in achieving the best results in next item prediction and learning users' representations.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Exploratory Analysismentioning

confidence: 99%

Variational User Modeling with Slow and Fast Features

Fazelnia

Simon

Anderson

et al. 2022

Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining

Self Cite

View full text Add to dashboard Cite

show abstract

“…In our work, we use the last method to approach sequential group recommendations. Hansen et al (2020) proposes another session-aware system for music recommendations. It uses a neural network architecture that models users' preferences as a sequence of embeddings, one for each session, suggesting that the user's recent selections and the session-level contextual variables (such as time and device used) are enough to predict the tracks a user will listen to.…”

Section: Sequential Recommendationsmentioning

confidence: 99%

Sequential group recommendations based on satisfaction and disagreement scores

et al. 2021

View full text Add to dashboard Cite

Recently, group recommendations have gained much attention. Nevertheless, most approaches consider only one round of recommendations. However, in a real-life scenario, it is expected that the history of previous recommendations is exploited to tailor the recommendations towards meeting the needs of the group members. Such history should include not only which items the system suggested, but also the reaction of the members to these items. This work introduces the problem of sequential group recommendations, by exploiting the concept of satisfaction and disagreement. Satisfaction describes how well the group received the suggested items. Disagreement describes the satisfaction bias among the group members. We utilize these concepts in three new aggregation methods, SDAA, SIAA and Average+, designed to address the specific challenges introduced by sequential group recommendations. We experimentally show the effectiveness of our methods using big real datasets for both stable and ephemeral groups.

show abstract

“…For a fair comparison they are therefore ignored when answering RQ2. Time of The Day: We define 5 time windows, where numbers in brackets correspond to the hour range: night (0-5), morning (6)(7)(8)(9)(10)(11), afternoon (12)(13)(14)(15)(16)(17), evening (18)(19)(20)(21)(22)(23), all (0-23). If a session spans across two hours, we round up and consider the whole session as either part of start or end hour.…”

Section: Experimental Settingsmentioning

confidence: 99%

“…Modelling and understanding skipping behaviour in music listening sessions arguably plays a crucial role in understanding user behaviour in modern streaming services. For instance, the skipping signal has already been used as a measure in heuristic-based playlist generation systems [9,25], user satisfaction [16,28], relevance [17], and as counterfactual estimators in Recommender Systems [22].…”

Section: Introductionmentioning

confidence: 99%

On Skipping Behaviour Types in Music Streaming Sessions

Meggetto

Revie

Levine

et al. 2021

Proceedings of the 30th ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

The ability to skip songs is a core feature in modern online streaming services. Its introduction has led to a new music listening paradigm and has changed the way users interact with the underlying services. Thus, understanding their skipping activity during listening sessions has acquired considerable importance. This is because such implicit feedback signal can be considered a measure of users' satisfaction (dissatisfaction or lack of interest), affecting their engagement with the platforms. Prior work has mainly focused on analysing the skipping activity at an individual song level. In this work, we investigate different behaviours during entire listening sessions with regards to the users' session-based skipping activity. To this end, we propose a data transformation and clustering-based approach to identify and categorise skipping types. Experimental results on the real-world music streaming dataset (Spotify) indicate four main types of session skipping behaviour. A subsequent analysis of short, medium, and long listening sessions demonstrate that these session skipping types are consistent across sessions of varying length. Furthermore, we discuss their distributional differences under various listening context information, i.e. day types (i.e. weekday and weekend), times of the day, and playlist types. CCS CONCEPTS• Information systems → Recommender systems.

show abstract

Contextual and Sequential User Embeddings for Large-Scale Music Recommendation

Cited by 86 publications

References 12 publications

Variational User Modeling with Slow and Fast Features

Variational User Modeling with Slow and Fast Features

Sequential group recommendations based on satisfaction and disagreement scores

On Skipping Behaviour Types in Music Streaming Sessions

Contact Info

Product

Resources

About