Two-stage Model for Automatic Playlist Continuation at Scale

  title={Two-stage Model for Automatic Playlist Continuation at Scale},
  author={Maksims Volkovs and Himanshu Rai and Zhaoyue Cheng and Ga Wu and Y. Lu and Scott Sanner},
  journal={Proceedings of the ACM Recommender Systems Challenge 2018},
Automatic playlist continuation is a prominent problem in music recommendation. Significant portion of music consumption is now done online through playlists and playlist-like online radio stations. Manually compiling playlists for consumers is a highly time consuming task that is difficult to do at scale given the diversity of tastes and the large amount of musical content available. Consequently, automated playlist continuation has received increasing attention recently [1, 7, 11]. The 2018… 

Figures and Tables from this paper

An Analysis of Approaches Taken in the ACM RecSys Challenge 2018 for Automatic Music Playlist Continuation

The ACM Recommender Systems Challenge 2018 focused on the task of automatic music playlist continuation, which is a form of the more general task of sequential recommendation, to recommend up to 500 tracks that fit the target characteristics of the original playlist.

Representation, Exploration and Recommendation of Music Playlists

This work 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, and compares the model with two other encoding architectures for baseline comparison.

Alleviating the cold-start playlist continuation in music recommendation using latent semantic indexing

A multi-stage retrieval system utilizing user-generated titles to alleviate the cold-start problem in automatic playlist continuation and outperforms the state-of-the-art approaches and improves recommendation accuracy significantly in three primary evaluation metrics.

Does Track Sequence in User-generated Playlists Matter?

This paper investigates 704,166 user-generated playlists of a major music streaming provider and studies the consistency of a variety of audio features and metadata between subsequent tracks in playlists, and relates this variance to the corresponding variance computed on a position-independent set of tracks.

Combining Objective Measures of Playlist Quality to Evaluate Automatically Generated Playlists

A new evaluation metric that aims to optimise playlist quality based on user preference, bridging the gap between human and machine evaluation is proposed and discussed.

Global-Local Similarity Function for Automatic Playlist Generation

Object and subjective evaluations show that GLSF-based APG achieves better smooth transition and ensure the long-term content consistency among the tracks, and a promising way to improve APG algorithms is proposed.

Using Latent Semantics of Playlist Titles and Descriptions to Enhance Music Recommendations

This work clusters 20,065 playlists with both titles and descriptions into 562 groups using track vectors learned by word2vec model on over 1 million playlists, presenting a simple and promising solution to the cold-start problem in music recommendation.

Music Artist Classification with Convolutional Recurrent Neural Networks

The results improve upon baseline works, verify the influence of the producer effect on classification performance and demonstrate the trade-offs between audio length and training set size.

Evaluating Recommender System Algorithms for Generating Local Music Playlists

This paper alters the standard evaluation procedure such that the algorithms only rank tracks by local artists for each of the eight different cities, and finds that the neighborhood-based approach (IIN) performs best for long-tail local music recommendation.

Controllable Music Playlist Generation Based on Knowledge Graph and Reinforcement Learning

This study proposes a novel music playlist generation method based on a knowledge graph and reinforcement learning and introduces the following two new ideas: using the informative knowledge graph data to promote efficient optimization through reinforcement learning, and setting the flexible reward function that target users can design the parameters of itself to guide target users to new types of music tracks.



Recsys challenge 2018: automatic music playlist continuation

The ACM Recommender Systems Challenge 2018 focused on automatic music playlist continuation, which is a form of the more general task of sequential recommendation, to recommend up to 500 tracks that fit the target characteristics of the original playlist.

Automatic playlist generation based on tracking user’s listening habits

Two simple algorithms that track the listening habits and form a listener model—a profile of listening habits are developed and the listener model is then used for automatic playlist generation.

Large-scale user modeling with recurrent neural networks for music discovery on multiple time scales

This paper presents a new approach to model users through recurrent neural networks by sequentially processing consumed items, represented by any type of embeddings and other context features, and obtains semantically rich user representations, which capture a user's musical taste over time.

Steerable Playlist Generation by Learning Song Similarity from Radio Station Playlists

This paper demonstrates a method for learning song transition probabilities from audio features extracted from songs played in professional radio station playlists and is able to generate steerable playlists by choosing the next song to play not simply based on that prior, but on a tag cloud that the user is able of manipulate to express the high-level characteristics of the music he wishes to listen to.

Deep Neural Networks for YouTube Recommendations

This paper details a deep candidate generation model and then describes a separate deep ranking model and provides practical lessons and insights derived from designing, iterating and maintaining a massive recommendation system with enormous user-facing impact.

Language Modeling with Gated Convolutional Networks

A finite context approach through stacked convolutions, which can be more efficient since they allow parallelization over sequential tokens, is developed and is the first time a non-recurrent approach is competitive with strong recurrent models on these large scale language tasks.

Factorization meets the neighborhood: a multifaceted collaborative filtering model

The factor and neighborhood models can now be smoothly merged, thereby building a more accurate combined model and a new evaluation metric is suggested, which highlights the differences among methods, based on their performance at a top-K recommendation task.

An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling

A systematic evaluation of generic convolutional and recurrent architectures for sequence modeling concludes that the common association between sequence modeling and recurrent networks should be reconsidered, and convolutionals should be regarded as a natural starting point for sequence modeled tasks.

Collaborative Filtering for Implicit Feedback Datasets

This work identifies unique properties of implicit feedback datasets and proposes treating the data as indication of positive and negative preference associated with vastly varying confidence levels, which leads to a factor model which is especially tailored for implicit feedback recommenders.

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Applied to a state-of-the-art image classification model, Batch Normalization achieves the same accuracy with 14 times fewer training steps, and beats the original model by a significant margin.