AUT Journal of Modeling and Simulation

AUT Journal of Modeling and Simulation

A Hybrid Deep Graph and Kernel Ensemble Approach to Mental Health Prediction in Social Network

Document Type : Research Article

Authors
Nazila Taghvaei 1
Behrooz Masoumi 2
MohammadReza Keyvanpour 3
Omid Sojoodi 2
1 Faculty of Computer and Information Technology Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran.
2 Faculty of Computer and Information Technology Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran
3 Department of Computer Engineering, Alzahra University, Vanak, Tehran, Iran.
10.22060/miscj.2026.23944.5403
Abstract
Mental-health forecasting from social-media data is a complex multimodal challenge involving temporal, textual, and relational information. This study presents a hybrid two-stage framework that integrates a Long Short-Term Memory (LSTM)-based graph ensemble with an Ensemble Deep Kernel Learning (EDKL) meta-model to predict depressive risk and emotional trajectories within online social networks. In Stage 1, user-level representations are encoded using an LSTM encoder combined with multiple graph neural backbones, including Graph Convolutional Networks (GCN), Graph Attention Networks (GAT), and Graph Transformer Networks (GTN). Their outputs are stacked and calibrated via logistic regression to produce reliable depression-risk probabilities. In Stage 2, an EDKL meta-learner aggregates predictions from heterogeneous deep models (MLP, CNN, and LSTM) through kernel ridge regression with hybrid kernels optimized by a meta-heuristic search algorithm. This hybrid architecture supports robust, fine-grained forecasting across temporal, behavioral, and relational dimensions. Experiments on publicly available Twitter and MHASN datasets demonstrate substantial improvements over transformer-based and single-stage baselines, achieving up to 99% accuracy with consistently low error variance. The study also addresses ethical considerations related to privacy, bias, and potential misuse, emphasizes reproducibility through transparent experimental protocols, and outlines promising directions for future multimodal extensions, including richer linguistic, visual, and interaction signals for clinically relevant mental-health monitoring.
Keywords
Mental Health
Social Networks
LSTM
Graph Neural Networks
Ensemble Deep Kernel Learning
Depression Detection
Time-Series Forecasting
Subjects
1.         Jiao Y., Z.K., Wei X., Carlisle N.B., Keller C.J., Oathes D.J., Fonzo G.A., Zhang Y., Deep graph learning of multimodal brain networks defines treatment-predictive signatures in major depression. Molecular Psychiatry, 30 (9), 2025.
2.         Ahmed I., e.a., Explainable AI for depression detection and severity assessment from social media. JMIR Mental Health 2025.
3.         Tang Y., e.a., Anxiety disorder identification via subspace-enhanced hypergraph neural network (seHGNN). Neurocomputing, 2025.
4.         Yang J., L.Q., Wang J., Hybrid CNN–LSTM model for mental-health monitoring on social media. Computers in Human Behavior, 2023.
5.         Rahman M.M., A.K.M., et al., Hybrid deep neural ensemble for mental-health assessment from text. Expert Systems with Applications, 231, 2023.
6.         Chen C., J.X., Xu L., Zhao J., Multimodal graph neural networks for emotion recognition in online users. Information Fusion, 96, 2023.
7.         Hochreiter S., S.J., Long short-term memory,. Neural Computation, 9 (8), 1997.
8.         Lee K.S., H.B.J., Graph machine learning on hidden networks and mental-health conditions in the middle-aged and elderly. Psychiatry Investigation, 21 (12), 2024.
9.         Kipf T.N., W.M., Semi-supervised classification with graph convolutional networks,. ICLR, 2017.
10.       Veličković P., C.G., Casanova A., Romero A., Liò P., Bengio Y., Graph Attention Networks. ICLR, 2018.
11.       Yun S., J.M., Kim R., Kang J., Kim H.J., Graph Transformer Networks. NeurIPS, 2019.
12.       Li S., Z.J., Xu Z., Explainable hybrid attention networks for emotion-aware social graph modeling. Knowledge-Based Systems, 2023.
13.       Xu Y., W.X., Affective computing through multi-level graph ensembles. IEEE Transactions on Affective Computing, 2023.
14.       Xing T., D.Y., Chen X., Zhou J., Xie X., Peng S., Adaptive multi-graph neural network with multimodal feature fusion for depression detection. Scientific Reports, 2024.
15.       Wang Z., L.D., Liu C., Temporal graph transformers for dynamic relational learning. Pattern Recognition, 2024.
16.       Qiu J., L.J., Fan X., Zhang M., Prior-guided adaptive time-frequency graph neural network for EEG depression diagnosis (ELPG-DTFS). arXiv:2509.24860,, 2025.
17.       OpenAI, G.-T.R., https://arxiv.org/abs/2303.08774. 2024.
18.       Google DeepMind, G., 2024., Model Overview and Technical Capabilities. 2024: https://arxiv.org/abs/2403.05530.
19.       T. Touvron, L.M., K. Stone, et al., LLaMA-2, Open Foundation and Fine-Tuned Chat Models. 2023: https://arxiv.org/abs/2307.09288.
20.       Y. Jiang, K.Z., M. De Choudhury,, Zero-Shot Depression Detection Using Large Language Models. 2024: https://arxiv.org/abs/2404.01345.
21.       Botteghi N., G.M., Brune C., Deep kernel learning of dynamical models from high-dimensional noisy data,. Scientific Reports, 12, 2022.
22.       Achituve I., e.a., Guided Deep Kernel Learning. ICML, 2023.
23.       Liu X., W.H., et al., Deep ensemble kernel fusion for adaptive emotion classification. Neural Networks, 2023.
24.       Zhang Q., e.a., A multimodal hybrid deep kernel network for social emotion prediction. Expert Systems with Applications, 2024.
25.       Zhang J., L.H., Sun Y., et al., Deep graph learning for social-media depression detection. IEEE Transactions on Affective Computing, 2022.
26.       Wu Z., P.S., Chen F., Long G., Zhang C., Yu P.S., , A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems, 32 (1) 2021.
27.       Bansal S., e.a., Deep kernel meta-learning for social signal processing. IEEE Access, 13, 2025.
28.       Ogbuanya, C.E., Adaora Obayi, Souad Larabi-Marie-Sainte, Amal O. Saad, and Lamia Berriche. , A hybrid optimization approach for accelerated multimodal medical image fusion. PLoS One 20, no. 7, 2025.
29.       Lin R., F.T., et al., Deep ensemble kernel learning for multimodal affective analysis. Knowledge-Based Systems, 2025.
30.       Zhang Y., e.a., Evaluating multimodal fusion strategies for social emotion prediction. Neurocomputing, 2025.
31.       Safa R., E.S.A., Sorourkhah A., predicting mental health using social media: A roadmap. arXiv:2301.10453, 2023.
32.       Ibrahimov Y., A.T., Yuan T., Mutallimov T., Hasanov E., AttentionDep: Domain-aware attention for explainable depression severity assessment. arXiv:2510.00706, 2025.
33.       Wang S., L.Z., Tan Z., Li J., Rasero J., Zhang A., Agarwal C., Interpretable neuropsychiatric diagnosis via concept-guided graph neural networks (CONCEPTNEURO),. arXiv:2510.03351, 2025.
34.       Geng S., Z.W., Xie J., Wang R., Ram S., From Detection to Discovery: Closed-loop depression detection and knowledge expansion on social media. arXiv:2510.23626, 2025.
35.       Srivastava P., e.a., Ethical AI for social good: Responsible mental-health modeling. AI Ethics, 4 (2), 2024.
36.       Putica, A., Rahul Khanna, Wiliam Bosl, Sudeep Saraf, and Juliet Edgcomb, Ethical decision-making for AI in mental health: the Integrated Ethical Approach for Computational Psychiatry (IEACP) framework. Psychological Medicine 55 2025.
37.       Hohman, F., Minsuk Kahng, Robert Pienta, and Duen Horng Chau, Visual analytics in deep learning: An interrogative survey for the next frontiers. IEEE Transactions on Visualization and Computer Graphics, 2018.
38.       Javeed A., A.P., Ghazi A.N., Saleem M.A., Berglund J.S., Predicting depression in older adults: A novel feature-selection and neural-network framework. Cognitive Computation, 2025.
39.       Koenig R., e.a., Human-centered explainable AI for psychological well-being applications. Nature Human Behaviour, 2025.
40.       Tahir, W.B., Shah Khalid, Sulaiman Almutairi, Mohammed Abohashrh, Sufyan Ali Memon, and Jawad Khan, Depression Detection in Social Media: A Comprehensive Review of Machine Learning and Deep Learning Techniques. IEEE Access, 2025.
41.       X. Wang, S.L., , EMNLP, LLM-Based Emotion Understanding in social media: A Comprehensive Benchmark. 2023: https://arxiv.org/abs/2311.05823.
AUT Journal of Modeling and Simulation
Volume 57, Issue 2
December 2025
Pages 173-190