Search

Your search keyword '"Ji, Jinchao"' showing total 8 results
Start Over Author "Ji, Jinchao" Publication Type Periodicals Publisher elsevier b.v.
8 results on '"Ji, Jinchao"'

5. Relationship-aware contrastive learning for social recommendations.

Author

Ji, Jinchao, Zhang, Bingjie, Yu, Junchao, Zhang, Xudong, Qiu, Dinghang, and Zhang, Bangzuo

Subjects
*RECOMMENDER systems, *SUPERVISED learning, *SOCIAL interaction
Abstract

The popularity bias is an outstanding challenge in recommendation systems. Prevalent work based on contrastive learning (CL) alleviates this issue but neglects the relationship among data, which limits the ability of CL and leads to a loss of personalized features of users/items, and thus degrades the performance of the recommendation systems. In this paper, we propose a model-agnostic contrastive learning framework, called Relationship-aware Contrastive Learning (ReACL), to make recommendations to users. ReACL framework utilizes the relationship homophily among data to achieve a uniform distribution for node and align the relevant features to preserve personalized features and avoids the problem of popularity bias. We first combine a graph of user-item interaction with that of a social network to mine views on the relationships of users on social networks as well as those of item commonality. We then apply these two views to construct contrastive learning pairs on the user and item sides, respectively. We design two sample selectors and an augmented contrastive loss function to preserve personalized features. Furthermore, we jointly optimize the tasks of supervised learning and contrastive learning. Finally, we conduct extensive experiments on four public datasets, and the results show that the proposed ReACL framework outperforms state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

8. Optimization analysis of droplet dust removal mass on tilted superhydrophobic surface based on regulating PV tilt angle and drop height.

Author

Zhang, Dong, Yan, Chengtao, Bai, Jianhua, Yu, Kai, Zhang, Jingrui, Ji, Jinchao, and An, Zhoujian

Subjects
*DUST removal, *SUPERHYDROPHOBIC surfaces, *DUST, *HYDROPHOBIC surfaces, *ANGLES, *LIGHT transmission
Abstract

• The motor behavior of droplet on tilted superhydrophobic dust surface was observed. • A dynamic model of droplet rolling dust removal was established. • The droplet cleaning effect can be controlled by the drop height and tilt angle. This study proposes a method for adjusting the mass of droplet dust removal based on droplet height and photovoltaic tilt angle, utilizing a model of droplet motion on inclined superhydrophobic surfaces. A droplet's motion on dusty and inclined hydrophobic surfaces is analyzed using a high-speed digital imaging system. The impact of drop height and photovoltaic tilt angle on the mass of dust removed by the droplet is experimentally investigated. The findings reveal that the mass of droplet dust removal increases quickly with the rise of drop height and tilt angle. The modification of the tilt angle yields a more pronounced enhancement in dust removal efficiency. At an angle of 40°, the drop height rises from 0.5 cm to 10 cm, leading to an extra 6.6 mg of dust removal per drop. At a height of 0 cm, the mass increased by 10.62 mg after raising the angle from 10° to 40°. Droplet impact on the surface is prone to shattering. The emitted minute droplets will adhere to the surface and form a cohesive ash band, with a visible light transmission rate of merely 67 %. When a droplet rolls on a tilted superhydrophobic surface, there's a dust removal threshold. Once exceeded, the rolling shape tends to become ellipsoidal. The experimental and model predictions of droplet dust removal mass are of the same order of magnitude and differ by a few micrograms. Therefore, the theoretical model can be used to improve the efficiency of droplet dust removal by adjusting the drop height and photovoltaic angle. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results