Abstract:Multimedia content is dominating today's Web information. The nature of multimedia user-item interactions is 1/0 binary implicit feedback (e.g., photo likes, video views, song downloads, etc.), which can be collected at a larger scale with a much lower cost than explicit feedback (e.g., product ratings). However, the majority of existing collaborative filtering (CF) systems are not well-designed for multimedia recommendation, since they ignore the implicitness in users' interactions with multimedia content. We argue that, in multimedia recommendation, there exists item- and component-level implicitness which blurs the underlying users' preferences. The item-level implicitness means that users' preferences on items (e.g. photos, videos, songs, etc.) are unknown, while the component-level implicitness means that inside each item users' preferences on different components (e.g. regions in an image, frames of a video, etc.) are unknown. For example, a 'view'' on a video does not provide any specific information about how the user likes the video (i.e.item-level) and which parts of the video the user is interested in (i.e.component-level). In this paper, we introduce a novel attention mechanism in CF to address the challenging item- and component-level implicit feedback in multimedia recommendation, dubbed Attentive Collaborative Filtering (ACF). Specifically, our attention model is a neural network that consists of two attention modules: the component-level attention module, starting from any content feature extraction network (e.g. CNN for images/videos), which learns to select informative components of multimedia items, and the item-level attention module, which learns to score the item preferences. ACF can be seamlessly incorporated into classic CF models with implicit feedback, such as BPR and SVD++, and efficiently trained using SGD. Through extensive experiments on two real-world multimedia Web services: Vine and Pinterest, we show that ACF significantly outperforms state-of-the-art CF methods.

ACCF: Learning Attentional Conformity for Collaborative Filtering

Using latent class models for neighbors selection in collaborative filtering

A unified latent factor correction scheme for collaborative filtering

Improving Recommendations With Collaborative Factors

Confidence-Learning Based Collaborative Filtering With Heterogeneous Implicit Feedbacks

Discrete limited attentional collaborative filtering for fast social recommendation

BCFNet: A Balanced Collaborative Filtering Network with Attention Mechanism

Deep Social Collaborative Filtering

A Community-Based Collaborative Filtering Method for Social Recommender Systems

Online Collaborative Filtering With Implicit Feedback

Collaborative Filtering Based on Diffusion Models: Unveiling the Potential of High-Order Connectivity

Attentive Collaborative Filtering: Multimedia Recommendation with Item- and Component-Level Attention

Collaborative Filtering with Topic and Social Latent Factors Incorporating Implicit Feedback

Listwise Collaborative Filtering

Contextual Attention Model for Social Recommendation.

Confluence: conformity influence in large social networks.

NtCF: Neural Trust-Aware Collaborative Filtering Toward Hierarchical Recommendation Services

Ranking-Oriented Collaborative Filtering

Hypergraph Contrastive Collaborative Filtering

Attention-Based Adaptive Model To Unify Warm And Cold Starts Recommendation

Kernel-CF: Collaborative filtering done right with social network analysis and kernel smoothing