Context-aware music recommendation based on latenttopic sequential patterns

Hariri, Negar; Mobasher, Bamshad; Burke, Robin

doi:10.1145/2365952.2365979

Cited by 238 publications

(166 citation statements)

References 7 publications

Supporting

Mentioning

165

Contrasting

Unclassified

Order By: Relevance

“…Park et al [26] presented a context-aware music recommender, which utilized several kinds of context information, including noise, light level, weather, and time. Hariri et al [13] adopted an LDA model to infer the topic probability distribution of songs with tags and discovered a pattern of topics in the song sequences, which can be used as contexts to improve the performance of music recommendation.…”

Section: Environment-related Context Based Approachesmentioning

confidence: 99%

“…In fact, the contexts may not be captured with a static set of factors, but rather, it is dynamic and can be inferred from users' interactions with the system. More specifically, the contexts are reflected in the sequences of music pieces played or liked by the users in their current interactions with the system [13], such as recent playlists, so it is feasible to infer the contextual information from the users' listening behaviors. Furthermore, users' historical listening records indicate lots of information, such as the features of music pieces and the users' preferences for music, and a music recommender system should be able to infer the user's contexts and musical preferences from the given music pieces liked or listened to by her/him and recommend appropriate music pieces to satisfy her/his real-time requirements.…”

Section: Motivationmentioning

confidence: 99%

“…Observation 3: every user's contextual musical preferences usually maintain stable within a period (usually one kind of taste), which are reflected in their recent listening records [13].…”

Section: Motivationmentioning

confidence: 99%

“…For 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 example, none of these four music pieces (13)(14)(15)(16) are similar with the other music pieces (1-12) in Table 8. Thirdly, some slight differences in styles and genres of music pieces are also shown by the learned embeddings, which shows that the learned embeddings by MEM can effectively capture the accurate features of the corresponding music pieces.…”

Section: Illustrations Of Selected Music Pieces' Embeddingsmentioning

confidence: 99%

“…More specifically, contexts are reflected in the sequence of music pieces liked or listened to by the user in her/his current interaction with the system, such as recent playlists [13]. For example, in Pandora 2 (an online music streaming service website), users create different playlists by choosing different track seeds or artists.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Learning to embed music and metadata for context-aware music recommendation

et al. 2017

View full text Add to dashboard Cite

Contextual factors greatly influence users' musical preferences, so they are beneficial remarkably to music recommendation and retrieval tasks. However, it still needs to be studied how to obtain and utilize the contextual information. In this paper, we propose a context-aware music recommendation approach, which can recommend music pieces appropriate for users' contextual preferences for music. In analogy to matrix factorization methods for collaborative filtering, the proposed approach does not require music pieces to be represented by features ahead, but it can learn the representations from users' historical listening records. Specifically, the proposed approach first learns music pieces' embeddings (feature vectors in low-dimension continuous space) from music listening records and corresponding metadata. Then it infers and models users' global and contextual preferences for music from their listening records with the learned embeddings. Finally, it recommends appropriate music pieces according to the target user's preferences to satisfy her/his real-time requirements. Experimental evaluations on a real-world dataset show that the proposed approach outperforms baseline methods in terms of precision, recall, F1 score, and hitrate. Especially, our approach has better performance on sparse datasets.

show abstract

Section: Environment-related Context Based Approachesmentioning

confidence: 99%

Section: Motivationmentioning

confidence: 99%

“…Observation 3: every user's contextual musical preferences usually maintain stable within a period (usually one kind of taste), which are reflected in their recent listening records [13].…”

Section: Motivationmentioning

confidence: 99%

Section: Illustrations Of Selected Music Pieces' Embeddingsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Learning to embed music and metadata for context-aware music recommendation

et al. 2017

View full text Add to dashboard Cite

show abstract

Building effective recommender systems for tourists

Massimo

Ricci

2022

AI Magazine

View full text Add to dashboard Cite

Recommender systems (RSs) are personalized information search and discovery applications helping users to identify and choose useful items and information. In this paper, we focus on the tourism application scenario and its specific requirements. We discuss a novel RS approach that copes with the specific application constraints of the domain and produces recommendations that better match the true needs of tourists. We illustrate the proposed next POI recommendation approach in a case study and we compare it with a state-of-the-art nearest neighbor-based next item RS. With the analysis of this case study, we aim at illustrating the specific features of the compared approaches also with the goal to raise the discussion on RSs validation methods, with a particular attention to tourism applications. We finally discuss some significant limitations of current evaluation approaches that must be addressed in future studies.

show abstract