On an artist’s profile page, music streaming services frequently recommend a ranked list of ”similar artists” that fans also liked. However, implementing such a feature is challenging for new artists, for which usage data on the service (e.g. streams or likes) is not yet available. In this paper, we model this cold start similar artists ranking problem as a link prediction task in a directed and attributed graph, connecting artists to their top-k most similar neighbors and incorporating side musical information. Then, we leverage a graph autoencoder architecture to learn node embedding representations from this graph, and to automatically rank the top-k most similar neighbors of new artists using a gravity-inspired mechanism. We empirically show the flexibility and the effectiveness of our framework, by addressing a real-world cold start similar artists ranking problem on a global music streaming service. Along with this paper, we also publicly release our source code and the industrial data from our experiments.
Description
Cold Start Similar Artists Ranking with Gravity-Inspired Graph Autoencoders | Fifteenth ACM Conference on Recommender Systems
%0 Conference Paper
%1 salhagalvan2021start
%A Salha-Galvan, Guillaume
%A Hennequin, Romain
%A Chapus, Benjamin
%A Tran, Viet-Anh
%A Vazirgiannis, Michalis
%B Fifteenth Conference on Recommender Systems
%D 2021
%I ACM
%K autoencoder ranking recommender
%R 10.1145/3460231.3474252
%T Cold Start Similar Artists Ranking with Gravity-Inspired Graph Autoencoders
%U https://doi.org/10.1145%2F3460231.3474252
%X On an artist’s profile page, music streaming services frequently recommend a ranked list of ”similar artists” that fans also liked. However, implementing such a feature is challenging for new artists, for which usage data on the service (e.g. streams or likes) is not yet available. In this paper, we model this cold start similar artists ranking problem as a link prediction task in a directed and attributed graph, connecting artists to their top-k most similar neighbors and incorporating side musical information. Then, we leverage a graph autoencoder architecture to learn node embedding representations from this graph, and to automatically rank the top-k most similar neighbors of new artists using a gravity-inspired mechanism. We empirically show the flexibility and the effectiveness of our framework, by addressing a real-world cold start similar artists ranking problem on a global music streaming service. Along with this paper, we also publicly release our source code and the industrial data from our experiments.
@inproceedings{salhagalvan2021start,
abstract = {On an artist’s profile page, music streaming services frequently recommend a ranked list of ”similar artists” that fans also liked. However, implementing such a feature is challenging for new artists, for which usage data on the service (e.g. streams or likes) is not yet available. In this paper, we model this cold start similar artists ranking problem as a link prediction task in a directed and attributed graph, connecting artists to their top-k most similar neighbors and incorporating side musical information. Then, we leverage a graph autoencoder architecture to learn node embedding representations from this graph, and to automatically rank the top-k most similar neighbors of new artists using a gravity-inspired mechanism. We empirically show the flexibility and the effectiveness of our framework, by addressing a real-world cold start similar artists ranking problem on a global music streaming service. Along with this paper, we also publicly release our source code and the industrial data from our experiments.},
added-at = {2021-09-30T08:42:57.000+0200},
author = {Salha-Galvan, Guillaume and Hennequin, Romain and Chapus, Benjamin and Tran, Viet-Anh and Vazirgiannis, Michalis},
biburl = {https://www.bibsonomy.org/bibtex/20643ceeec84b422841a4a2458d47d5e1/jaeschke},
booktitle = {Fifteenth Conference on Recommender Systems},
description = {Cold Start Similar Artists Ranking with Gravity-Inspired Graph Autoencoders | Fifteenth ACM Conference on Recommender Systems},
doi = {10.1145/3460231.3474252},
interhash = {e894da4140d9fecb1554803c359c9dd3},
intrahash = {0643ceeec84b422841a4a2458d47d5e1},
keywords = {autoencoder ranking recommender},
month = sep,
publisher = {ACM},
timestamp = {2021-09-30T08:42:57.000+0200},
title = {Cold Start Similar Artists Ranking with Gravity-Inspired Graph Autoencoders},
url = {https://doi.org/10.1145%2F3460231.3474252},
year = 2021
}