Exploration in Interactive Personalized Music Recommendation

Xin-xi, Wang; Wang, Yi; Hsu, David; Wang, Ye

doi:10.1145/2623372

Cited by 80 publications

(51 citation statements)

References 41 publications

(53 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…What's more, the advanced recommendation approaches start to apply reinforcement learning [8] to the recommendation process, and consider the recommendation task as a decision problem to provide more accurate recommendations. Wang et al [11] proposed a reinforcement learning framework based on Bayesian model to balance the exploration and exploitation of users' preferences for recommendation. To learn user preferences, it uses a Bayesian model that accounts for both audio content and the novelty of recommendations.…”

Section: Related Workmentioning

confidence: 99%

A Hybrid Recommendation for Music Based on Reinforcement Learning

Wang

2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

The key to personalized recommendation system is the prediction of users' preferences. 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 process which can capture the minor changes of listeners' preferences sensitively. 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. We firstly use weighted matrix factorization (WMF) and convolutional neural network (CNN) to learn and extract the song feature vectors. In order to capture the changes of listeners' preferences sensitively, we innovatively enhance simulating interaction process of listeners and update the model continuously based on their preferences both for songs and song transitions. The extensive experiments on real-world datasets validate the effectiveness of the proposed PHRR on song sequence recommendation compared with the state-of-the-art recommendation approaches.

show abstract

Section: Related Workmentioning

confidence: 99%

A Hybrid Recommendation for Music Based on Reinforcement Learning

Wang

2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

show abstract

“…Exploitation-Exploration of items is also a hot topic in RS eld [5,10,17,19,22,24,35,43]. Exploring more items can introduce more diversity in recommendation results.…”

Section: Related Workmentioning

confidence: 99%

Beyond Personalization

Xiao

Zhao

Pan

et al. 2019

Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery &Amp; Data Mining

View full text Add to dashboard Cite

An e ective content recommendation in modern social media platforms should bene t both creators to bring genuine bene ts to them and consumers to help them get really interesting content. In this paper, we propose a model called Social Explorative A ention Network (SEAN) for content recommendation. SEAN uses a personalized content recommendation model to encourage personal interests driven recommendation. Moreover, SEAN allows the personalization factors to a end to users' higher-order friends on the social network to improve the accuracy and diversity of recommendation results. Constructing two datasets from a popular decentralized content distribution platform, Steemit, we compare SEAN with state-of-the-art CF and content based recommendation approaches. Experimental results demonstrate the e ectiveness of SEAN in terms of both Gini coe cients for recommendation equality and F1 scores for recommendation performance.

show abstract

“…2.2.1 Contextual Multi-Armed Bandit models. A group of work [5,7,23,40,44,50] begin to formulate the problem as a Contextual Multi-Armed Bandit (MAB) problem, where the context contains user and item features. [23] assumes the expected reward is a linear function of the context.…”

Section: Reinforcement Learning In Recommendationmentioning

confidence: 99%

DRN

Zheng

Zhang

Zheng

et al. 2018

Proceedings of the 2018 World Wide Web Conference on World Wide Web - WWW '18

447

View full text Add to dashboard Cite

In this paper, we propose a novel Deep Reinforcement Learning framework for news recommendation. Online personalized news recommendation is a highly challenging problem due to the dynamic nature of news features and user preferences. Although some online recommendation models have been proposed to address the dynamic nature of news recommendation, these methods have three major issues. First, they only try to model current reward (e.g., Click Through Rate). Second, very few studies consider to use user feedback other than click / no click labels (e.g., how frequent user returns) to help improve recommendation. Third, these methods tend to keep recommending similar news to users, which may cause users to get bored. Therefore, to address the aforementioned challenges, we propose a Deep Q-Learning based recommendation framework, which can model future reward explicitly. We further consider user return pattern as a supplement to click / no click label in order to capture more user feedback information. In addition, an effective exploration strategy is incorporated to find new attractive news for users. Extensive experiments are conducted on the offline dataset and online production environment of a commercial news recommendation application and have shown the superior performance of our methods.

show abstract

Exploration in Interactive Personalized Music Recommendation

Cited by 80 publications

References 41 publications

A Hybrid Recommendation for Music Based on Reinforcement Learning

A Hybrid Recommendation for Music Based on Reinforcement Learning

Beyond Personalization

DRN

Contact Info

Product

Resources

About