Abstract:User counterparts, such as user attributes in social networks or user interests, are the keys to more natural Human–Computer Interaction (HCI). In addition, users’ attributes and social structures help us understand the complex interactions in HCI. Most previous studies have been based on supervised learning to improve the performance of HCI. However, in the real world, owing to signal malfunctions in user devices, large amounts of abnormal information, unlabeled data, and unsupervised approaches (e.g., the clustering method) based on mining user attributes are particularly crucial. This paper focuses on improving the clustering performance of users’ attributes in HCI and proposes a deep graph embedding network with feature and structure similarity (called DGENFS) to cluster users’ attributes in HCI applications based on feature and structure similarity. The DGENFS model consists of a Feature Graph Autoencoder (FGA) module, a Structure Graph Attention Network (SGAT) module, and a Dual Self-supervision (DSS) module. First, we design an attributed graph clustering method to divide users into clusters by making full use of their attributes. To take full advantage of the information of human feature space, a k-neighbor graph is generated as a feature graph based on the similarity between human features. Then, the FGA and SGAT modules are utilized to extract the representations of human features and topological space, respectively. Next, an attention mechanism is further developed to learn the importance weights of different representations to effectively integrate human features and social structures. Finally, to learn cluster-friendly features, the DSS module unifies and integrates the features learned from the FGA and SGAT modules. DSS explores the high-confidence cluster assignment as a soft label to guide the optimization of the entire network. Extensive experiments are conducted on five real-world data sets on user attribute clustering. The experimental results demonstrate that the proposed DGENFS model achieves the most advanced performance compared with nine competitive baselines.

Co-clustering Interactions Via Attentive Hypergraph Neural Network

CoHomo: A Cluster-Attribute Correlation Aware Graph Clustering Framework

Hyperedge Interaction-aware Hypergraph Neural Network

Hierarchical attention network for attributed community detection of joint representation

A Deep Graph Network with Multiple Similarity for User Clustering in Human–Computer Interaction

A Versatile Framework for Attributed Network Clustering via K-Nearest Neighbor Augmentation

Attention-driven Graph Clustering Network.

Multi-Behavior Enhanced Heterogeneous Graph Convolutional Networks Recommendation Algorithm based on Feature-Interaction

Adaptive Harmony Learning and Optimization for Attributed Graph Clustering

GraphHINGE: Learning Interaction Models of Structured Neighborhood on Heterogeneous Information Network

A Novel Clustering Algorithm Based on Multi-Layer Features and Graph Attention Networks

Dynamical Attention Hypergraph Convolutional Network for Group Activity Recognition

Cross Multi-Type Objects Clustering in Attributed Heterogeneous Information Network

Integrated Heterogeneous Graph Attention Network for Incomplete Multi-modal Clustering

InteractNet: Social Interaction Recognition for Semantic-rich Videos

Shared-Attribute Multi-Graph Clustering with Global Self-Attention

Robust Clustering Model Based on Attention Mechanism and Graph Convolutional Network

Homophily-Related: Adaptive Hybrid Graph Filter for Multi-View Graph Clustering

Towards attributed graph clustering using enhanced graph and reconstructed graph structure

Cascaded Attention: Adaptive and Gated Graph Attention Network for Multiagent Reinforcement Learning

Multilayer Graph Contrastive Clustering Network