Filters








1,276 Hits in 4.7 sec

Automatic, Personalized, and Flexible Playlist Generation using Reinforcement Learning

Shun-Yao Shih, Heng-Yu Chi
2018 Zenodo  
Considering a playlist as a sequence of words, we first train our attention RNN language model on baseline recommended playlists.  ...  By exploiting the techniques of deep learning and reinforcement learning, in this paper, we consider music playlist generation as a language modeling problem and solve it by the proposed attention language  ...  Therefore, it will be great if we can refine playlists based on different preferences of users on the fly.  ... 
doi:10.5281/zenodo.1492370 fatcat:w3fajmvqpbalrc62ubwi3djnge

Automatic, Personalized, and Flexible Playlist Generation using Reinforcement Learning [article]

Shun-Yao Shih, Heng-Yu Chi
2018 arXiv   pre-print
Considering a playlist as a sequence of words, we first train our attention RNN language model on baseline recommended playlists.  ...  By exploiting the techniques of deep learning and reinforcement learning, in this paper, we consider music playlist generation as a language modeling problem and solve it by the proposed attention language  ...  Therefore, it will be great if we can refine playlists based on different preferences of users on the fly.  ... 
arXiv:1809.04214v1 fatcat:m5aen2fqpjeidngoixuqg6uyy4

DJ-MC: A Reinforcement-Learning Agent for Music Playlist Recommendation [article]

Elad Liebman, Maytal Saar-Tsechansky, Peter Stone
2015 arXiv   pre-print
In this work we present DJ-MC, a novel reinforcement-learning framework for music recommendation that does not recommend songs individually but rather song sequences, or playlists, based on a model of  ...  In recent years, there has been growing focus on the study of automated recommender systems.  ...  DJ-MC In this section we introduce DJ-MC, a novel reinforcement learning approach to a playlist-oriented, personalized music recommendation system.  ... 
arXiv:1401.1880v2 fatcat:powiz6oe7rdhvo6ihul3f2kmce

Controllable Music Playlist Generation Based on Knowledge Graph and Reinforcement Learning

Keigo Sakurai, Ren Togo, Takahiro Ogawa, Miki Haseyama
2022 Sensors  
In this study, we propose a novel music playlist generation method based on a knowledge graph and reinforcement learning.  ...  The playlist generation is one of the most important multimedia techniques, which aims to recommend music tracks by sensing the vast amount of musical data and the users' listening histories from music  ...  reinforcement learning.  ... 
doi:10.3390/s22103722 pmid:35632130 pmcid:PMC9144078 fatcat:2hxfod6vdfbgpgah44getnt4rq

Collective Noise Contrastive Estimation for Policy Transfer Learning

Weinan Zhang, Ulrich Paquet, Katja Hofmann
2016 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
We demonstrate the effectiveness of our approach by learning an effective policy for an online radio station jointly from user-generated playlists, and usage data collected in an exploration bucket.  ...  We address the problem of learning behaviour policies to optimise online metrics from heterogeneous usage data.  ...  Acknowledgements We sincerely thank Thore Graepel and Noam Koenigstein for many helpful discussions, and Nir Nice and the Microsoft Recommendations Team for supporting this project.  ... 
doi:10.1609/aaai.v30i1.10153 fatcat:c6lp6cjdhjbf7fb4aawyzhbndq

Efficient Online Learning to Rank for Sequential Music Recommendation

Pedro Dalla Vecchia Chaves, Bruno L. Pereira, Rodrygo L. T. Santos
2022 Proceedings of the ACM Web Conference 2022  
One notable component of these services are playlists, which can be dynamically generated in a sequential manner based on the user's feedback during a listening session.  ...  To overcome these limitations, we propose a novel online learning to rank approach which efficiently explores the space of candidate recommendation models by restricting itself to the orthogonal complement  ...  Reinforcement learning (RL) based algorithms can be employed to continuously adapt the online model, being a flexible alternative to improve the song recommendation task by leveraging users' feedback  ... 
doi:10.1145/3485447.3512116 fatcat:2tjnrrvzwfgzvflqdts35xhtmq

Representation, Exploration and Recommendation of Music Playlists [article]

Piyush Papreja and Hemanth Venkateswara and Sethuraman Panchanathan
2019 arXiv   pre-print
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.  ...  Playlists have become a significant part of our listening experience because of the digital cloud-based services such as Spotify, Pandora, Apple Music.  ...  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  ... 
arXiv:1907.01098v1 fatcat:44bd7ojegjhg3mwma2xjhyweha

Recommending Music Curators: A Neural Style-Aware Approach [chapter]

Jianling Wang, James Caverlee
2020 Lecture Notes in Computer Science  
Three unique features of the proposed framework are: (i) models of curation style to capture the coverage of music and curator's individual style in assigning tracks to playlists; (ii) a curation-based  ...  We propose a framework for personalized music curator recommendation to connect users with curators who have matching curation style.  ...  For example, [12] predicts the next track based on a listener's preferences and most-recently played tracks. DJ-MC [14] aims to recommend track sequences based on reinforcement learning.  ... 
doi:10.1007/978-3-030-45439-5_13 fatcat:osea2pkvefbido24k5s2awy3tq

A Hybrid Recommendation for Music Based on Reinforcement Learning [chapter]

Yu Wang
2020 Lecture Notes in Computer Science  
In this paper, we propose a personalized hybrid recommendation algorithm for music based on reinforcement learning (PHRR) to recommend song sequences that match listeners' preferences better.  ...  However, almost all existing music recommendation approaches only learn listeners' preferences based on their historical records or explicit feedback, without considering the simulation of interaction  ...  In this paper, we propose a personalized hybrid recommendation algorithm for music based on reinforcement learning (PHRR).  ... 
doi:10.1007/978-3-030-47426-3_8 fatcat:iwaof2fgyzfbrl6lhda4zjemzm

Embedding Factorization Models for Jointly Recommending Items and User Generated Lists

Da Cao, Liqiang Nie, Xiangnan He, Xiaochi Wei, Shunzhi Zhu, Tat-Seng Chua
2017 Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval - SIGIR '17  
However, li le a ention has been paid to recommend user generated lists (e.g., playlists and booklists).  ...  On one hand, user generated lists contain rich signal about item co-occurrence, as items within a list are usually gathered based on a speci c theme.  ...  Based on these criteria, we ultimately obtained 18, 528 users, 123, 628 songs, 22, 864 playlists, 1, 128, 065 user-song interactions, and 528, 128 user-playlist interactions.  ... 
doi:10.1145/3077136.3080779 dblp:conf/sigir/CaoN0WZC17 fatcat:kpfuipswajhyxbykoac7jhgvgq

Diversifying Music Recommendations [article]

Houssam Nassif, Kemal Oral Cansizlar, Mitchell Goodman, SVN Vishwanathan
2018 arXiv   pre-print
We compare submodular and Jaccard methods to diversify Amazon Music recommendations. Submodularity significantly improves recommendation quality and user engagement.  ...  Based on minutes streamed, both diversity measures fare better than baseline. This result reinforces the notion that diversity affects recommendation quality (Zhang et al., 2012) .  ...  Attribution: Nassif, Cansizlar, Goodman, Vishwanathan, "Diversifying Music Recommendations" Machine Learning for Music Discovery Workshop at the 33 rd International Conference on Machine Learning, New  ... 
arXiv:1810.01482v1 fatcat:4vxiutoemvgc3gqz56cuolltyq

"LOVING OR LOATHING LYNDA" A PILOT STUDY INVESTIGATION INTO THE INTEGRATION OF VIDEO E-LEARNING RESOURCES WITHIN AN UNDERGRADUATE LEISURE MANAGEMENT DEGREE UNIT

Lynsey Lynsey
2016 PEOPLE International Journal of Social Sciences  
Based on these two recommendations and further informed by previous research within the last decade on the impact of video tutorials/online e-learning resources in HE, such as De Vaney (2009), Njenga and  ...  Two recommendations were proposed on the need to research 'student attitudes to different methods of e-learning' and 'the impact of e-learning in relation to assessment'.  ...  Recommendations The following recommendations reflect the key outcomes from the review of literature on e-learning and subsequent analysis of Lynda.com monitoring data and primary data from the pilot study  ... 
doi:10.20319/pijss.2016.s21.3175 fatcat:e7jcqbn4cvctpfafcr7ipkqtku

Datafication and the push for ubiquitous listening in music streaming

Rasmus Rex Pedersen
2020 MedieKultur: Journal of Media and Communication Research  
This article discusses Spotify's approach to music recommendation as dataficationof listening. It discusses the hybrid types of music recommendation that Spotifypresents to users.  ...  The article explores how datafication is connected to Spotify'spush for the personalization and contextualization of music recommendationsbased on a combination of the cultural knowledge found in editorial  ...  (Joven, 2018 and genres might suggest a dichotomy between playlists curated based on the stylistic features of the musical content and playlists curated based on the intended listening situations.  ... 
doi:10.7146/mediekultur.v36i69.121216 fatcat:jpptdkhbh5cn3drm22vj33c5ea

Metrics and decisions-making in music streaming

Arnt Maasø, Anja Nylund Hagen
2019 Popular Communication  
When discussing these findings, we draw attention to the reinforcing feedback loops between metrics, data-based decisions and algorithms, questioning whether datafication acts to intensify trending events  ...  The analysis concludes that they rely on a growing volume of data when making decisions about what to promote, and how.  ...  Notes on contributors Arnt Maasø (Dr. art.) is Associate Professor of media and communication at the University of Oslo.  ... 
doi:10.1080/15405702.2019.1701675 fatcat:2brpx3n3nves3oxsbgstfim23q

Exploring acoustic similarity for novel music recommendation

Derek S Cheng, Thorsten Joachims, Douglas Turnbull
2020 Zenodo  
Most commercial music services rely on collaborative filtering to recommend artists and songs.  ...  In this paper, we therefore seek to understand how content-based approaches can be used to more effectively recommend songs from these lesser known artists.  ...  [23] stress the importance of serendipity and warn about the dangers of self-reinforcing "filter bubbles" when music recommender systems focus too much on optimizing accuracy.  ... 
doi:10.5281/zenodo.4245500 fatcat:ds4234rfzngarexpylpvynonjm
« Previous Showing results 1 — 15 out of 1,276 results