Personalized Latent Structure Learning for Recommendation

Zhang, Shengyu; Feng, Fuli; Kuang, Kun; Zhang, Wenqiao; Zhao, Zhou; Yang, Hongxia; Chua, Tat-Seng; Wu, Fei

doi:10.1109/tpami.2023.3247563

Cited by 8 publications

(3 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, content-based filtering and unattended learning models offer insightful analysis of specific playlist features. This allows for the provision of recommendations categorized by genre, artist, and even mixed suggestions [51,52]. The overarching objective of these personalized music recommendation systems is to elevate user satisfaction and optimize recommendation effectiveness, thus enriching the overall listening experience.…”

Section: G Personalizationmentioning

confidence: 99%

Beyond the Trends: Evolution and Future Directions in Music Recommender Systems Research

Amiri,

Shahverdi,

Haddadi

et al. 2024

IEEE Access

View full text Add to dashboard Cite

The study of Music Recommender Systems (MRS) has become a crucial field in the age of digital music consumption, influencing how people discover and interact with music. This comprehensive analysis examines the complex field of MRS research from 2005 to 2023, With the growing importance of music recommendation systems in enhancing user experience, it is crucial to comprehend their development. By utilizing rigorous social network analysis, statistical measures, and factor analysis, our investigation not only identifies important themes and influential contributors but also emphasizes the complex and diverse nature of MRS. The trend of the field exhibits a significant increase between 2017 and 2021, with periodic oscillations that highlight its dynamic nature. This analysis offers a broad perspective by examining highly cited articles, current researchers, and local sources. Factorial analysis uncovers thematic clusters, highlighting collaborative filtering, user experience, emotion identification, and reinforcement learning. A scientific mapping analysis classifies research themes in different historical periods, with a focus on important areas such as collaborative filtering, hybrid recommendation, sentiment analysis, and emotion identification. A review of thematic evolution highlights the importance of digitalization, emotion recognition, personalization, user experience, and collaborative filtering in determining the future directions of research. Although there has been a recent decrease in general interest, the investigation of context-aware models and hybrid techniques offers encouraging opportunities for further inquiry. This research enhances our comprehension of MRS dynamics, leading to future improvements and developments in the field. Ultimately, it improves the music discovery experience for people globally.

show abstract

Section: G Personalizationmentioning

confidence: 99%

Beyond the Trends: Evolution and Future Directions in Music Recommender Systems Research

Amiri,

Shahverdi,

Haddadi

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…And so on, the historical features saved in the 𝑘-th epoch will be used to calculate the orthogonal direction used in the 𝑘 + 1-th epoch. [6,39,40] shows that in the training of classification models, only 5 epochs are usually needed to make the shift of features small enough. And 𝑘 is significantly larger than 5, so there is no need to be concerned about the feature shift not being small enough.…”

Section: End-to-end Training With Ourmentioning

confidence: 99%

“…Long-tailed recognition is an important challenge in computer vision, manifested by models trained on long-tailed data that tend to perform poorly for classes with few samples. Previous research has attributed this phenomenon to the fact that the few samples in the tail classes do not well represent their true distribution, resulting in a shift between the test and training domains [4,31,40]. Numerous methods have been proposed to mitigate model bias.…”

Section: Introductionmentioning

confidence: 99%

Orthogonal Uncertainty Representation of Data Manifold for Robust Long-Tailed Learning

Ma,

Jiao,

Liu

et al. 2023

Proceedings of the 31st ACM International Conference on Multimedia

View full text Add to dashboard Cite

In scenarios with long-tailed distributions, the model's ability to identify tail classes is limited due to the under-representation of tail samples. Class rebalancing, information augmentation, and other techniques have been proposed to facilitate models to learn the potential distribution of tail classes. The disadvantage is that these methods generally pursue models with balanced class accuracy on the data manifold, while ignoring the ability of the model to resist interference. By constructing noisy data manifold, we found that the robustness of models trained on unbalanced data has a long-tail phenomenon. That is, even if the class accuracy is balanced on the data domain, it still has bias on the noisy data manifold. However, existing methods cannot effectively mitigate the above phenomenon, which makes the model vulnerable in long-tailed scenarios. In this work, we propose an Orthogonal Uncertainty Representation (OUR) of feature embedding and an end-to-end training strategy to improve the long-tail phenomenon of model robustness. As a general enhancement tool, OUR has excellent compatibility with other methods and does not require additional data generation, ensuring fast and efficient training. Comprehensive evaluations on long-tailed datasets show that our method significantly improves the long-tail phenomenon of robustness, bringing consistent performance gains to other long-tailed learning methods. CCS CONCEPTS• Computing methodologies → Computer vision; Image representations; Supervised learning by classification.

show abstract

End-to-End Optimization of Quantization-Based Structure Learning and Interventional Next-Item Recommendation

Fu,

Miao,

Zhang

et al. 2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Personalized Latent Structure Learning for Recommendation

Cited by 8 publications

References 46 publications

Beyond the Trends: Evolution and Future Directions in Music Recommender Systems Research

Beyond the Trends: Evolution and Future Directions in Music Recommender Systems Research

Orthogonal Uncertainty Representation of Data Manifold for Robust Long-Tailed Learning

End-to-End Optimization of Quantization-Based Structure Learning and Interventional Next-Item Recommendation

Contact Info

Product

Resources

About