Semantic manifold modularization-based ranking for image recommendation. (December 2021)

• Record Type: Journal Article
• Authors: Jian, Meng; Guo, Jingjing; Zhang, Chenlin; Jia, Ting; Wu, Lifang; Yang, Xun; Huo, Lina
• Is Part Of: Pattern recognition. Volume 120(2021)
• Publication Date: 2021-12
• Subjects: Manifold propagation -- Modularization -- Image recommendation -- User interest; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2021.108100
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: The proposed MMR employs visual correlations of images that users consumed to reveal and infer users' interests by interest propagation over the visual graph of images instead of propagating collaborative signals over users' sparse interaction graph. We constrain manifold learning within visual groups adaptively to propagate users' interests and prevent bias propagated across semantics as a tradeoff between personality and propagation smoothness. For image recommendation, the proposed MMR introduces manifold modularization to perform interest propagation in a decomposed manner and reduce computational burden exponentially. Abstract: As the Internet confronts the multimedia explosion, it becomes urgent to investigate personalized recommendation for alleviating information overload and improving users' experience. Most personalized recommendation approaches pay their attention to collaborative filtering over users' interactions, which suffers greatly from the highly sparse interactions. In image recommendation, visual correlations among images that users consumed provide a piece of intrinsic evidence to reveal users' interests. It inspires us to investigate image recommendation over the dense visual graph of images instead of the sparse user interaction graph. In this paper, we propose a semantic manifold modularization-based ranking (MMR) for image recommendation. MMR leverages the dense visual manifold to propagate users' historical records and infer user-image correlations for image recommendation. Especially, it constrains interest propagation within semantic visual compact groups by manifold modularization to make a tradeoff between users' personality and graph smoothness in propagation. Experimental results demonstrate that user-consumed visual correlations play actively to capture users' interests, and the proposed MMR can infer user-image correlations via visual manifold propagation for image recommendation.