Title:Graph Neural Networks for User Satisfaction Classification in Human-Computer Interaction
Abstract:This study focuses on the problem of user satisfaction classification and proposes a framework based on graph neural networks to address the limitations of traditional methods in handling complex interaction relationships and multidimensional features. User behaviors, interface elements, and their potential connections are abstracted into a graph structure, and joint modeling of nodes and edges is used to capture semantics and dependencies in the interaction process. Graph convolution and attention mechanisms are introduced to fuse local features and global context, and global pooling with a classification layer is applied to achieve …
Title:Graph Neural Networks for User Satisfaction Classification in Human-Computer Interaction
Abstract:This study focuses on the problem of user satisfaction classification and proposes a framework based on graph neural networks to address the limitations of traditional methods in handling complex interaction relationships and multidimensional features. User behaviors, interface elements, and their potential connections are abstracted into a graph structure, and joint modeling of nodes and edges is used to capture semantics and dependencies in the interaction process. Graph convolution and attention mechanisms are introduced to fuse local features and global context, and global pooling with a classification layer is applied to achieve automated satisfaction classification. The method extracts deep patterns from structured data and improves adaptability and robustness in multi-source heterogeneous and dynamic environments. To verify effectiveness, a public user satisfaction survey dataset from Kaggle is used, and results are compared with multiple baseline models across several performance metrics. Experiments show that the method outperforms existing approaches in accuracy, F1-Score, AUC, and Precision, demonstrating the advantage of graph-based modeling in satisfaction prediction tasks. The study not only enriches the theoretical framework of user modeling but also highlights its practical value in optimizing human-computer interaction experience.
| Subjects: | Human-Computer Interaction (cs.HC) |
| Cite as: | arXiv:2511.04166 [cs.HC] |
| (or arXiv:2511.04166v1 [cs.HC] for this version) | |
| https://doi.org/10.48550/arXiv.2511.04166 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Rui Liu [view email] [v1] Thu, 6 Nov 2025 08:11:20 UTC (306 KB)