Denoising-Guided Deep Reinforcement Learning For Social Recommendation

Du, Qihan; Yu, Li; Li, Huiyuan; Leng, Youfang; Ou, Ningrui; Xiang, Junyao

doi:10.1109/icassp43922.2022.9747580

Cited by 2 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, r𝑢𝑣 at the end of 𝑘th curriculum period, i.e., 𝑘𝐷th epoch, is smoothed by that of the last period. Mathematically, r𝑢𝑣 is updated as follows, (11) where a hyperparameter 𝛽 controls smoothness. On the other hand, Motivated by Dunbar's number theory suggesting an upper limit of a user's close friends [12], we propose to adaptively denoise the user 𝑢's social graph based on her friend number, i.e., |R 𝑢 |.…”

Section: Robust Denoising Of Social Networkmentioning

confidence: 99%

“…Early works use simple statistics like number of co-interactions [31] to indicate the confidence degree of social relations, which is generally ineffective due to data sparsity [40]. In order to capture observed diversity of social influence, attention mechanism [50,55], fine-grained contextual information [14], expectation-maximization method [47] or reinforcement learning [11] can be leveraged to learn adaptive weights among various friends. Although above techniques can be applied in prevalent GSocRec models, they generally fall short of learning the degree of social influence effectively due to a lack of groundtruth labels.…”

Section: Related Work 51 Graph Denoising For Social Recommendationmentioning

confidence: 99%

See 1 more Smart Citation

Robust Preference-Guided Denoising for Graph based Social Recommendation

Quan

Ding

Gao

et al. 2023

Proceedings of the ACM Web Conference 2023

View full text Add to dashboard Cite

Graph Neural Network (GNN) based social recommendation models improve the prediction accuracy of user preference by leveraging GNN in exploiting preference similarity contained in social relations. However, in terms of both effectiveness and efficiency of recommendation, a large portion of social relations can be redundant or even noisy, e.g., it is quite normal that friends share no preference in a certain domain. Existing models do not fully solve this problem of relation redundancy and noise, as they directly characterize social influence over the full social network. In this paper, we instead propose to improve graph based social recommendation by only retaining the informative social relations to ensure an efficient and effective influence diffusion, i.e., graph denoising. Our designed denoising method is preference-guided to model social relation confidence and benefits user preference learning in return by providing a denoised but more informative social graph for recommendation models. Moreover, to avoid interference of noisy social relations, it designs a self-correcting curriculum learning module and an adaptive denoising strategy, both favoring highly-confident samples. Experimental results on three public datasets demonstrate its consistent capability of improving two state-of-the-art social recommendation models by robustly removing 10-40% of original relations. We release the source code at https://github.com/tsinghua-fib-lab/Graph-Denoising-SocialRec.

show abstract

Section: Robust Denoising Of Social Networkmentioning

confidence: 99%