Filters








4,079 Hits in 4.2 sec

The Natural Language Of Playlists

Brian McFee, Gert R. G. Lanckriet
2011 Zenodo  
To demonstrate the technique, we developed a suite of simple baseline playlisters, and evaluated their performance on naturally occurring playlists.  ...  We have presented a simple, automatic evaluation procedure for playlist algorithms.  ...  The universe of human-generated playlists therefore constitutes a natural language, and playlisters are models of the language of playlists.  ... 
doi:10.5281/zenodo.1418119 fatcat:vddmegq4ovauxi2i7z4hpq67sq

Representation, Exploration and Recommendation of Music Playlists [article]

Piyush Papreja and Hemanth Venkateswara and Sethuraman Panchanathan
2019 arXiv   pre-print
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  ...  Owing to the meteoric rise in the usage of playlists, recommending playlists is crucial to music services today.  ...  Due to these similarities, there have been many works which employ techniques from the field of natural language processing by translating the problem to an already solved-problem in natural language processing  ... 
arXiv:1907.01098v1 fatcat:44bd7ojegjhg3mwma2xjhyweha

Music Playlist Title Generation: A Machine-Translation Approach [article]

SeungHeon Doh, Junwon Lee, Juhan Nam
2021 arXiv   pre-print
Considering the orderless nature of music tracks in a playlist, we propose two techniques that remove the order of the input sequence.  ...  We take a sequence of track IDs as input and a sequence of words in a playlist title as output, adapting the sequence-to-sequence framework based on Recurrent Neural Network (RNN) and Transformer to the  ...  In Melon, playlist titles are written in both Korean and English (some of titles are mixed with both languages).  ... 
arXiv:2110.07354v1 fatcat:5hisxz2c5nhplhgqfvvreere5m

Playlist prediction via metric embedding

Shuo Chen, Josh L. Moore, Douglas Turnbull, Thorsten Joachims
2012 Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '12  
An empirical evaluation shows that the LME is substantially more accurate than adaptations of smoothed n-gram models commonly used in natural language processing.  ...  The key goal of automated playlist generation is to provide the user with a coherent listening experience.  ...  Modeling playlists as a Markov chain connects to a large body of work on sequence modeling in natural language processing and speech recognition.  ... 
doi:10.1145/2339530.2339643 dblp:conf/kdd/ChenMTJ12 fatcat:66pwgzzqafakbhwo427pvstz4u

Enhancing Slot Tagging with Intent Features for Task Oriented Natural Language Understanding using BERT [article]

Shruthi Hariharan, Vignesh Kumar Krishnamurthy, Utkarsh, Jayantha Gowda Sarapanahalli
2022 arXiv   pre-print
We evaluate our techniques on benchmark spoken language datasets SNIPS and ATIS, as well as over a large private Bixby dataset and observe an improved slot-tagging performance over state-of-the-art models  ...  In this paper, we examine the effect of leveraging intent label features through 3 techniques in the slot tagging task of joint intent and slot detection models.  ...  These systems usually comprise multiple domains, and cater to executing one or more actions after a particular domain and intent are chosen based on natural language input from the user.  ... 
arXiv:2205.09732v2 fatcat:oo4lqxc3tff5pcilis7dtfetfu

When users generate music playlists: When words leave off, music begins?

Simone Stumpf, Sam Muscroft
2011 2011 IEEE International Conference on Multimedia and Expo  
We present the results of an explorative study that focused on the language of musically untrained end users for playlist choices, in a variety of listening contexts.  ...  Our results indicate that there are a number of opportunities for playlist recommendation or retrieval systems, particularly by taking context into account.  ...  ACKNOWLEDGMENTS We thank Charles Inskip and the participants of our study. This work was funded as part of the City University London pump-priming grant scheme.  ... 
doi:10.1109/icme.2011.6012152 dblp:conf/icmcs/StumpfM11 fatcat:qnjuwhlswbe4bclf3ivqvo5tby

Curating the Fieldwork Playlist

Kieran Fenby-Hulse
2019 Suomen Antropologi  
I argue that the playlist asks us to reflect on the field of anthropology and to consider the role of the voice, the body, the mind with anthropology, as well as the role digital technologies, ethics,  ...  The chapter examines this history of the digital playlist before going on to analyse the varied musical contributions from curatorial, musicological, and anthropological perspetives.  ...  Indeed, the array of languages and styles within the playlist can foster feelings of otherness and unfamiliarity.  ... 
doi:10.30676/jfas.v43i2.77751 fatcat:yvhiwye66zhk5jos552ulnilja

Hypergraph Models Of Playlist Dialects

Brian McFee, Gert R. G. Lanckriet
2012 Zenodo  
This viewpoint, borrowed from the statistical natural language processing literature, enables the automatic evaluation and optimization of a model by computing the likelihood of it generating examples  ...  language, we propose to model playlists as a collection of dialects, each of which may exhibit its own particular structure.  ... 
doi:10.5281/zenodo.1415618 fatcat:ecmoyau3gjeybbsduihyz4vnii

Music Playlist Recommendation via Preference Embedding

Chih-Ming Chen, Chun-Yao Yang, Chih-Chun Hsia, Yian Chen, Ming-Feng Tsai
2016 ACM Conference on Recommender Systems  
After the embedding process, we then use the learned representations to perform the task of playlist recommendation.  ...  In the light of this, this paper proposes a preference embedding based on a user-song-playlist graph to learn the preference representations of these three entities.  ...  In the field of natural language processing, the techniques are usually referred to as word embedding for language modeling and feature learning to map words or phrases into a low-dimensional vector space  ... 
dblp:conf/recsys/ChenYHCT16 fatcat:dcmgqy7omvhtldqqkvqyirwqsi

Quick Lists: Enriched Playlist Embeddings for Future Playlist Recommendation [article]

Brett Vintch
2020 arXiv   pre-print
Recommending playlists to users in the context of a digital music service is a difficult task because a playlist is often more than the mere sum of its parts.  ...  The embeddings also capture information about playlist sequencing, and are enriched with side information about the playlist user.  ...  There are strong analogies between the task of representing playlists and that of representing natural language.  ... 
arXiv:2006.12382v1 fatcat:zlsn4t3porhell7epgmehwwedu

Learning To Embed Songs And Tags For Playlist Prediction

Joshua L. Moore, Shuo Chen, Thorsten Joachims, Douglas Turnbull
2012 Zenodo  
Another interesting direction for future work is the modeling of long-range dependencies in playlists.  ...  Such long-range dependencies could capture the amount of redundancy/repetition that a user may seek, versus how much a playlist provides variety and explores new music.  ...  Modeling playlists as a Markov chain connects to a large body of work on sequence modeling in natural language processing (NLP) and speech recognition.  ... 
doi:10.5281/zenodo.1416966 fatcat:akli7c4dpfbhng2ztqqkxl5voe

A Language-Based Approach to Variability Analysis

Bruno Santana da Silva, Simone Diniz Junqueira Barbosa, Julio Cesar Sampaio do Prado Leite
2008 Workshop em Engenharia de Requisitos  
In this paper, we explore the variability of the user's domain language to promote the requirements engineers' reflection on the need to deal with variations and what strategies are adequate to deal with  ...  the user goals on user interface language.  ...  The first author also thanks FAPERJ for the financial support to his Doctoral program.  ... 
dblp:conf/wer/SilvaBL08 fatcat:mefulpenwjfzvjmzb3pn7ugfku

Autoencoders for Next-Track-Recommendation

Michael Vötter, Eva Zangerle, Maximilian Mayerl, Günther Specht
2019 Workshop Grundlagen von Datenbanken  
This work investigates the suitability and applicability of autoencoders for the task of playlist continuation. We utilize autoencoders and hence, representation learning to continue playlists.  ...  Our approach is inspired by the usage of autoencoders to denoise images and we consider the playlist without the missing next-track as a noisy input.  ...  They come up with the idea to consider playlist generation as a natural language modeling problem instead of an information retrieval problem.  ... 
dblp:conf/gvd/VotterZMS19 fatcat:7zik2kdbbbgt7fhezzf2ofu2au

Towards Music Captioning: Generating Music Playlist Descriptions [article]

Keunwoo Choi and George Fazekas and Brian McFee and Kyunghyun Cho and Mark Sandler
2017 arXiv   pre-print
In the proposed method, audio content analysis and natural language processing are adopted to utilise the information of each track.  ...  Recommending automatically generated music playlists (e.g. personalised playlists) introduces the problem of generating descriptions.  ...  One common method is to add descriptions of a music item or playlist, e.g.  ... 
arXiv:1608.04868v2 fatcat:knitoad3g5bodbcijow3hwk5zi

Understanding Music Playlists [article]

Keunwoo Choi, George Fazekas, Mark Sandler
2015 arXiv   pre-print
As music streaming services dominate the music industry, the playlist is becoming an increasingly crucial element of music consumption.  ...  Better understanding of the playlist is there- fore necessary for developing better playlist gen- eration algorithms.  ...  Therefore, it is natural that the focus of music recommendation problem is shifting towards the playlist generation problem. However, the understanding of playlists is still shallow.  ... 
arXiv:1511.07004v1 fatcat:jfdwijaaovcqvjvofycpx3pyzq
« Previous Showing results 1 — 15 out of 4,079 results