International Journal of Modern Education and Computer Science (IJMECS)
IJMECS Vol. 17, No. 6, 8 Dec. 2025
Cover page and Table of Contents: PDF (size: 878KB)
NeSy-Guidance: A Neuro-Symbolic Knowledge Graph for Academic Recommendations Combining Rule-Based Reasoning and Neural Inference
PDF (878KB), PP.48-64
Views: 0 Downloads: 0
Author(s)
Zineb Elkaimbillah
1,*
Zineb Mcharfi
1
Mohamed Khoual
1
Bouchra El Asri
1
Index Terms
Neuro-Symbolic Approach, Knowledge Graph, Graph Convolutional Network (GCN), Symbolic Rule Reasoning, Policy-alignment, Academic Recommendation
Abstract
The growing complexity of academic and career decision-making requires intelligent systems that can deliver personalized recommendations while ensuring strict compliance with institutional policies and incorporating evolving contextual factors. This paper introduces NeSy-Guidance, a neuro-symbolic recommendation approach that combines symbolic rule reasoning with graph-based neural inference over a dynamic and time-aware academic Knowledge Graph (KG). The model encodes students, programs, and contextual entities including regions and emerging trends while integrating regulatory constraints derived from Moroccan admission policies. It applies a two-stage reasoning pipeline: a symbolic layer enforcing hard eligibility rules and soft preference-based adjustments mined automatically from the knowledge graph, and a Graph Convolutional Network (GCN) layer trained with a weighted loss to address class imbalance and capture latent student–program compatibility. A weighted score fusion mechanism integrates both inference outputs, achieving a balance between interpretability, adaptability, and predictive performance. Evaluated on a real-world dataset of 800 students and 325 academic programs, NeSy-Guidance outperforms three state-of-the-art baselines in both accuracy and policy compliance. It achieves 83.8% accuracy, 74.5% precision@5, 75.4% F1-score, and ensures 100% compliance with institutional eligibility rules. Furthermore, a qualitative survey confirmed positive student satisfaction regarding the clarity and relevance of recommendations. These results demonstrate the effectiveness of hybrid reasoning and validate NeSy-Guidance as a reliable, explainable, and regulation-aware academic guidance system capable of adapting to regional disparities, emerging academic trends, and evolving student preferences.
Cite This Paper
Zineb Elkaimbillah, Zineb Mcharfi, Mohamed Khoual, Bouchra El Asri, "NeSy-Guidance: A Neuro-Symbolic Knowledge Graph for Academic Recommendations Combining Rule-Based Reasoning and Neural Inference", International Journal of Modern Education and Computer Science(IJMECS), Vol.17, No.6, pp. 48-64, 2025. DOI:10.5815/ijmecs.2025.06.04
Reference
[1]Allman, B., Kimmons, R., Wang, W. et al. Trends and Topics in Educational Technology, 2024 Edition. TechTrends 68, 402–410 (2024). https://doi.org/10.1007/s11528-024-00950-5
[2]Zhou, Y., Shirazi, S. Factors Influencing Young People’s STEM Career Aspirations and Career Choices: A Systematic Literature Review. Int J of Sci and Math Educ (2025). https://doi.org/10.1007/s10763-025-10552-z
[3]Wang, S., Wang, Y., Sivrikaya, F. et al. Data science for next-generation recommender systems. Int J Data Sci Anal 16, 135–145 (2023). https://doi.org/10.1007/s41060-023-00404-w
[4]Torkashvand, A., Jameii, S.M. & Reza, A. Deep learning-based collaborative filtering recommender systems: a comprehensive and systematic review. Neural Comput & Applic 35, 24783–24827 (2023). https://doi.org/10.1007/s00521-023-08958-3
[5]Nair, A.M., Benny, O., George, J. (2021). Content Based Scientific Article Recommendation System Using Deep Learning Technique. In: Suma, V., Chen, J.IZ., Baig, Z., Wang, H. (eds) Inventive Systems and Control. Lecture Notes in Networks and Systems, vol 204. Springer, Singapore. https://doi.org/10.1007/978-981-16-1395-1_70
[6]Murgod, T.R., Reddy, P.S., Gaddam, S. et al. A Survey on Graph Neural Networks and its Applications in Various Domains. SN COMPUT. SCI. 6, 26 (2025). https://doi.org/10.1007/s42979-024-03543-4
[7]Dobson, J.E. On reading and interpreting black box deep neural networks. Int J Digit Humanities 5, 431–449 (2023). https://doi.org/10.1007/s42803-023-00075-w
[8]Peng, C., Xia, F., Naseriparsa, M. et al. Knowledge Graphs: Opportunities and Challenges. Artif Intell Rev 56, 13071–13102 (2023). https://doi.org/10.1007/s10462-023-10465-9
[9]Wang, X., He, X., Cao, Y., Liu, M., & Chua, T. S. (2019, July). Kgat: Knowledge graph attention network for recommendation. In Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining (pp. 950-958), https://dl.acm.org/doi/10.1145/3292500.3330989.
[10]Ma, W., Zhang, M., Cao, Y., Jin, W., Wang, C., Liu, Y., Ma, S., & Ren, X. (2019). Jointly Learning Explainable Rules for Recommendation with Knowledge Graph. In The World Wide Web Conference (Pp. 1210-1221), 1210‑1221. https://doi.org/10.1145/3308558.3313607
[11]Chen, X., Tang, T., Ren, J., Lee, I., Chen, H., & Xia, F. (2021, December). Heterogeneous graph learning for explainable recommendation over academic networks. In IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology (pp. 29-36). https://doi.org/10.1145/3498851.3498926
[12]Abu-Rasheed, H., Jumbo, C., Amin, R. A., Weber, C., Wiese, V., Obermaisser, R., & Fathi, M. (2025). LLM-Assisted Knowledge Graph Completion for Curriculum and Domain Modelling in Personalized Higher Education Recommendations. arXiv preprint arXiv:2501.12300.
[13]Colelough, B. C., & Regli, W. (2025). Neuro-Symbolic AI in 2024: A Systematic Review. arXiv preprint arXiv:2501.05435.
[14]Spillo, G., Musto, C., de Gemmis, M., Lops, P., & Semeraro, G. (2024). Recommender systems based on neuro-symbolic knowledge graph embeddings encoding first-order logic rules. User Modeling and User-Adapted Interaction, 34, 2039–2083. https://doi.org/10.1007/s11257-024-09417-x
[15]Serafini, L., & Garcez, A. D. A. (2016). Logic tensor networks: Deep learning and logical reasoning from data and knowledge. arXiv preprint arXiv:1606.04422.
[16]Chen, B., Peng, W., Wu, M., Zheng, B., & Zhu, S. (2023). Neural-symbolic recommendation with graph-enhanced information. arXiv preprint arXiv:2307.05036. https://arxiv.org/abs/2307.05036
[17]Xian, Y., Zhang, K., Wang, M., & Wang, Y. (2020). Neural-symbolic reasoning over knowledge graph for multi-stage explainable recommendation. arXiv preprint arXiv:2007.13207. https://arxiv.org/abs/2007.13207
[18]Zhang, Y., & Wang, Y. (2023). PSYCHIC: A neuro-symbolic framework for knowledge graph question-answering grounding. arXiv preprint arXiv:2310.12638. https://arxiv.org/abs/2310.12638
[19]Zeng, Z.; Cheng, Q.; Si, Y. Logical Rule-Based Knowledge Graph Reasoning: A Comprehensive Survey. Mathematics 2023, 11, 4486. https://doi.org/10.3390/math11214486
[20]Galárraga, L., Teflioudi, C., Hose, K. et al. Fast rule mining in ontological knowledge bases with AMIE+. The VLDB Journal 24, 707–730 (2015). https://doi.org/10.1007/s00778-015-0394-1N Ahmad et al. Students' Performance Prediction using Artificial Neural Network, IOP Conf. Ser.: Mater. Sci. Eng. 1176 012020, 2021
[21]Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., & Yu, P. S. (2019). A Comprehensive Survey on Graph Neural Networks. ArXiv. https://doi.org/10.1109/TNNLS.2020.2978386
[22]Zhang, M., & Chen, Y. (2018). Link prediction based on graph neural networks. Advances in neural information processing systems, 31
[23]Zineb, E., Zineb, M., Mohamed, K., & Bouchra, E. (2025). Building knowledge graph for relevant degree recommendations using semantic similarity search and named entity recognition. Indonesian Journal of Electrical Engineering and Computer Science, 37(1), 463-474. doi: http://doi.org/10.11591/ijeecs.v37.i1.pp463-474
[24]Alwateer, M., Atlam, E. S., Abd El-Raouf, M. M., Ghoneim, O. A., & Gad, I. (2024). Missing data imputation: A comprehensive review. Journal of Computer and Communications, 12(11), 53-75.