Abstract:Multi-view multi-label (MVML) learning is a framework for solving the problem of associating a single instance with a set of class labels in the presence of multiple types of data features. The extraction of shared features among multiple views for label prediction is a common MVML learning method. However, previous approaches assumed that the number and association degree of shared features were the same across views. In fact, they differ in the number and degree of association. The above assumption can lead to a poor communicability of the views. Therefore, this paper proposes an MVML learning method based on the inconsistent shared features extracted by the graph attention model. The first step is to extract the shared and private features of multiple views. Next, the graph attention mechanism is adopted to learn the association degree of shared features of different views and calculate the adjacency matrix and attention coefficient. The number of associations is determined by taking the obtained adjacency matrix as a mask matrix, while the association degree of shared features is measured by the attention weight matrix. Finally, the new shared features are obtained for multi-label prediction. We conducted experiments on seven MVML datasets to compare the proposed algorithm with seven advanced algorithms. The experimental results demonstrate the advantages of our algorithm.

What problem does this paper attempt to address?

The paper primarily addresses the issue of shared feature inconsistency in Multi-View Multi-Label (MVML) learning. Specifically, it tackles the problem of assuming the same quantity and degree of association for shared features across different views when, in reality, these features may vary in quantity and association degree across different views. The main contributions of the paper are as follows: 1. **Proposes an MVML learning method based on Graph Attention Mechanism (GAT)**, which can handle the inconsistency of extracted shared features. 2. **Utilizes Graph Attention Networks (GAT) to learn the degree of association of shared features across different views**. By calculating the adjacency matrix and attention coefficients, and further determining the number of associations and measuring the degree of association between shared features, new shared features are obtained. 3. **Conducts extensive experimental validation**, comparing the proposed method with seven advanced algorithms on seven MVML datasets. The experiments demonstrate the effectiveness and superiority of the proposed method. The structure of the paper is arranged as follows: - The first section introduces the research background, namely the basic framework of multi-label learning and its application in real-world problems. - The second section reviews related work, including the current state of research on multi-view multi-label learning and graph attention networks. - The third section describes in detail the proposed model structure and implementation methods, including the extraction of shared and private features, the processing of shared features based on the graph attention mechanism, and the final label prediction. - The fourth section reports detailed experimental results and analysis, verifying the effectiveness of the proposed method. - The fifth section summarizes the entire paper. - The sixth section discusses the application directions and future research of multi-view multi-label classification. In summary, this paper aims to improve the shortcomings of existing MVML learning methods regarding shared feature extraction by introducing a graph attention mechanism to better capture the correlation of features across different views, thereby enhancing the performance of multi-view multi-label learning tasks.

Multi-view Multi-label Learning with Shared Features Inconsistency

Multi-View Clustering Via Simultaneously Learning Shared Subspace And Affinity Matrix

Multi-View Correlated Feature Learning by Uncovering Shared Component.

Beyond Shared Subspace: A View-Specific Fusion for Multi-View Multi-Label Learning.

Non-Aligned Multi-View Multi-Label Classification Via Learning View-Specific Labels

Multi-View Partial Multi-Label Learning with Graph-Based Disambiguation.

Consistent and Specific Multi-View Multi-Label Learning with Correlation Information

Multi-view multi-label learning for label-specific features via GLocal Shared Subspace Learning

L-VSM: Label-Driven View-Specific Fusion for Multiview Multilabel Classification

Multi-view multi-label learning for image annotation

Label driven latent subspace learning for multi-view multi-label classification

Embedded Feature Fusion for Multi-View Multi-Label Feature Selection

Label-Specific Alignment with Adversarial Multi-View Representation

Graph-based Multi-view Partial Multi-label Learning

Relation-Aware Alignment Attention Network for Multi-view Multi-label Learning.

Multi-View Partial Multi-Label Learning Via Graph-Fusion-Based Label Enhancement

Deep dual incomplete multi-view multi-label classification via label semantic-guided contrastive learning

A Unified Framework for Graph-Based Multi-View Partial Multi-Label Learning.

Consistency and Diversity Neural Network Multi-View Multi-Label Learning

Label recovery and label correlation co-learning for multi-view multi-label classification with incomplete labels

Multiview Multi-Instance Multilabel Active Learning