Graph Neural Networks to Detect Suspicious Patterns in Nigerian Elections

The 2023 Nigerian Presidential Election generated verifiable and compelling allegations of polling unit-level result manipulation in many states in Nigeria. This study presents a Graph Neural Network (GNN) approach to detect that manipulation automatically because the central observation is that electoral suspicious patterns in Nigeria does not happen at random or isolated polling units. It happens in clusters, because ward collation officers can alter or replace multiple result sheets under their authority simultaneously. That coordinated, geographically clustered nature of suspicious patterns is exactly what GNNs are built to exploit. This study models each polling unit as a node in a graph where edges connect units that share the same ward. A 2-layer GATv2Conv model (available in Pytorch) then learns suspicious patterns patterns across ward neighbourhoods rather than individual units in isolation. Labels are engineered from three signals: statistical anomaly thresholds, Isolation Forest pre-screening, and critically, forensic annotations from the Centre for Collaborative Investigative Journalism (CCIJ) as gold labels. Training on Lagos (11,911 polling units) and Rivers (4,769 polling units) produces Area Under the Precision Recall Curve, AUC-PR of 0.9648 and 0.9196 respectively, representing 12.2x and 6.5x lifts over random baselines. An ablation study shows that adding ward membership edges alone accounts for a 132% improvement in AUC-PR over a feature-only MLP baseline. Ward-level suspicious patterns maps produced by the model independently converges on the specific wards named in CCIJ's forensic reporting, including Rumueme Ward 7A in Obio/Akpor at mean suspicious patterns probability of 0.998.

Graph Neural Networks, Election Suspicious patterns Detection, GATv2Conv, Nigeria, Polling Unit Analysis, Semi-Supervised Learning

1. Brody, S., Alon, U., and Yahav, E. (2022). How attentive are graph attention networks? International Conference on Learning Representations (ICLR 2022). arXiv:2105.14491. [Google Scholar] [Crossref]

2. Centre for Collaborative Investigative Journalism and Daily Trust (2024). INEC's Broken Promises: A forensic investigation into Nigeria's 2023 election results. VEZA. https://veza.news/article/2025/03/31/broken-promises-of-transparency [Google Scholar] [Crossref]

3. Dou, Y., Liu, Z., Sun, L., Deng, Y., Peng, H., and Yu, P. S. (2020). Enhancing graph neural network-based suspicious patterns detectors against camouflaged suspicious patternssters. Proceedings of the 29th ACM International Conference on Information and Knowledge Management, pp. 315-324. [Google Scholar] [Crossref]

4. Fey, M. and Lenssen, J. E. (2019). Fast graph representation learning with PyTorch Geometric. ICLR Workshop on Representation Learning on Graphs and Manifolds. [Google Scholar] [Crossref]

5. Kipf, T. N. and Welling, M. (2017). Semi-supervised classification with graph convolutional networks. International Conference on Learning Representations (ICLR 2017). [Google Scholar] [Crossref]

6. Klimek, P., Yegorov, Y., Hanel, R., and Thurner, S. (2012). Statistical detection of systematic election irregularities. Proceedings of the National Academy of Sciences, 109(41), pp. 16469-16473. [Google Scholar] [Crossref]

7. Lin, T.-Y., Goyal, P., Girshick, R., He, K., and Dollar, P. (2017). Focal loss for dense object detection. IEEE International Conference on Computer Vision (ICCV 2017). [Google Scholar] [Crossref]

8. Liu, F., Zhao, Z., and Chen, H. (2024). Financial suspicious patterns detection using graph neural networks: A systematic review. Expert Systems with Applications, 240, 122156. [Google Scholar] [Crossref]

9. Liu, Y., Ao, X., Qin, Z., Chi, J., Feng, J., Yang, H., and He, Q. (2021). Pick and choose: A GNN-based imbalanced learning approach for suspicious patterns detection. The Web Conference (WWW 2021), pp. 3168-3177. [Google Scholar] [Crossref]

10. Mchenryspagg (2023). Outlier detection in election data. GitHub. https://github.com/mchenryspagg/Outlier-Detection-in-Election-Data [Google Scholar] [Crossref]

11. Mebane, W. R. (2008). Election forensics: The second-digit Benford's law test and recent American presidential elections. In Election Suspicious patterns: Detecting and Deterring Electoral Manipulation. Brookings Institution Press. [Google Scholar] [Crossref]

12. Paszke, A., Gross, S., Massa, F., and Lerer, A. et al. (2019). PyTorch: An imperative style, high-performance deep learning library. Advances in Neural Information Processing Systems 32. [Google Scholar] [Crossref]

13. Velickovic, P., Cucurull, G., Casanova, A., Romero, A., Lio, P., and Bengio, Y. (2018). Graph attention networks. International Conference on Learning Representations (ICLR 2018). [Google Scholar] [Crossref]

14. Ajilore, O. O., Eludire, A., Olumoye, M. Y., Adegunwa, O., Omotayo, O., Adekunle, A., ... & Olumoye, Y. (2023). Image classification algorithms for early detection of learning disabilities using visual data. International Journal of Innovative Science and Research Technology, 8(11), 815-819. [Google Scholar] [Crossref]

15. Ajilore, O. O., Phillip, A., Olumoye, M.Y., Eludire, A., Akanni, A., Adegunwa, O., (2025). Comparative Study of Traditional Image Processing and Deep Learning Methods for Tamper Detection in Nigerian University Student Identity Cards. International Journal of Computer Science and Information Security 19 (4), 109-126 [Google Scholar] [Crossref]

• Self-Supervised Learning for Oil Spill Detection in Synthetic Aperture Radar Imagery