Two-stage Model for Automatic Playlist Continuation at Scale

Volkovs, Maksims; Rai, Himanshu; Cheng, Zhaoyue; Wu, Ga; Lu, Yichao; Sanner, Scott

doi:10.1145/3267471.3267480

Cited by 33 publications

(30 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…8 We believe that our findings concerning the temporal relistening patterns of music genres (see Section 3.1) could help identify genres that users commonly listened to consecutively. We could then, for example, incorporate such genre sequences into the two-stage convolutional neural network (CNN) model for automatic playlist continuation that was proposed by Volkovs et al (2018). Finally, we would like to highlight that our approach could be easily leveraged by researchers and practitioners also for other related tasks (e.g., recommending music artists) and not only for genre prediction.…”

Section: Discussionmentioning

confidence: 99%

Modeling Popularity and Temporal Drift of Music Genre Preferences

Lex¹,

Kowald²,

Schedl

2020

Transactions of the International Society for Music Information Retrieval

View full text Add to dashboard Cite

In this paper, we address the problem of modeling and predicting the music genre preferences of users. We introduce a novel user modeling approach, BLL u , which takes into account the popularity of music genres as well as temporal drifts of user listening behavior. To model these two factors, BLL u adopts a psychological model that describes how humans access information in their memory. We evaluate our approach on a standard dataset of Last.fm listening histories, which contains fine-grained music genre information. To investigate performance for different types of users, we assign each user a mainstreaminess value that corresponds to the distance between the user's music genre preferences and the music genre preferences of the (Last.fm) mainstream. We adopt BLL u to model the listening habits and to predict the music genre preferences of three user groups: listeners of (i) niche, low-mainstream music, (ii) mainstream music, and (iii) medium-mainstream music that lies in-between. Our results show that BLL u provides the highest accuracy for predicting music genre preferences, compared to five baselines: (i) group-based modeling, (ii) user-based collaborative filtering, (iii) item-based collaborative filtering, (iv) frequency-based modeling, and (v) recencybased modeling. Besides, we achieve the most substantial accuracy improvements for the low-mainstream group. We believe that our findings provide valuable insights into the design of music recommender systems.

show abstract

Section: Discussionmentioning

confidence: 99%

Modeling Popularity and Temporal Drift of Music Genre Preferences

Lex¹,

Kowald²,

Schedl

2020

Transactions of the International Society for Music Information Retrieval

View full text Add to dashboard Cite

show abstract

“…Current research pertaining to playlist is in the areas of automatic playlist generation [8], [9] [10], and continuation [11] [12] [13]. Multiple solutions have been proposed to address these problems, like reinforcement learning [14] and Recurrent Neural Network-based models [4] for playlist generation and playlist continuation tasks.…”

Section: Why Playlist Embeddings?mentioning

confidence: 99%

Representation, Exploration and Recommendation of Playlists

Papreja

Venkateswara

Panchanathan

2020

Machine Learning and Knowledge Discovery in Databases

View full text Add to dashboard Cite

Playlists have become a significant part of our listening experience because of the digital cloudbased services such as Spotify, Pandora, Apple Music. Owing to the meteoric rise in the usage of playlists, recommending playlists is crucial to music services today. Although there has been a lot of work done in playlist prediction, the area of playlist representation hasn't received that level of attention. Over the last few years, sequence-to-sequence models, especially in the field of natural language processing, have shown the effectiveness of learned embeddings in capturing the semantic characteristics of sequences. We can apply similar concepts to music to learn fixed length representations for playlists and use those representations for downstream tasks such as playlist discovery, browsing, and recommendation. In this work, we formulate the problem of learning a fixed-length playlist representation in an unsupervised manner, using Sequence-to-sequence (Seq2seq) models, interpreting playlists as sentences and songs as words. We compare our model with two other encoding architectures for baseline comparison. We evaluate our work using the suite of tasks commonly used for assessing sentence embeddings, along with a few additional tasks pertaining to music, and a recommendation task to study the traits captured by the playlist embeddings and their effectiveness for the purpose of music recommendation.

show abstract

“…Furthermore, in [30], prediction and matrix factorisation models are combined for playlist continuation problem for Spotify music listening platform. They re-ranked the candidate tracks based on the results of the prediction and recommendation models by calculating latent factors for playlist and tracks using a prediction model after training the model using matrix factorisation.…”

Section: B Re-ranking In Recommender Systemsmentioning

confidence: 99%

Improving Session-Based Recommendation Adopting Linear Regression-Based Re-ranking

Esmeli

Bader-El-Den

Abdullahi

et al. 2020

2020 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

Due to the increase in the importance of giving real-time recommendation to e-commerce users, session-based recommender systems become more popular. Session-based recommendation systems have the ability to adapt quickly to respond to changes in user interests and newly added items. The ranking is the core part of recommender systems regardless of recommender system type. Re-ranking is applied to recommender systems to have more personalised recommendations by considering contextawareness. In this paper, we proposed an approach to re-rank recommended items by using a linear regression model. In our approach, we use users' current session features and temporal features of recommended items to measure a user's interest level on a recommended item. We focus on having better recall and precision scores with fewer recommendations to able to prove the success of our re-ranking strategy. We conduct computational experiments on six real-world datasets and show that after applying re-ranking, we can get higher recall and precision scores. These results confirm that taking user interest level on an item in a session into account can improve the chance of getting correct items in top 5 recommendations.

show abstract

Two-stage Model for Automatic Playlist Continuation at Scale

Cited by 33 publications

References 12 publications

Modeling Popularity and Temporal Drift of Music Genre Preferences

Modeling Popularity and Temporal Drift of Music Genre Preferences

Representation, Exploration and Recommendation of Playlists

Improving Session-Based Recommendation Adopting Linear Regression-Based Re-ranking

Contact Info

Product

Resources

About