AI-Driven Personalized Learning Framework for Computing Fields Recommendations in Secondary and Pre-University Contexts

Students or learners, especially those at the secondary and pre-university levels, face challenges in identifying suitable computing fields for personalized educational resources towards their tertiary education specialization pathways. Traditional advising methods/approaches such as guidance from school counselors and teachers often lack personalization and practical relevance because their knowledge of the field is limited. This research aims to close this gap by proposing Computing Field Recommender (ComReF), an end-to-end Machine Learning framework that integrates student assessment and a profile built from demographic attributes and assessment performance, with semantic resource generation to deliver personalized, interpretable, and pedagogically grounded recommendations. A Random Forest-based multi-class classification strategy predicts suitable computing fields and corresponding proficiency levels, with empirical results showing improved performance over Decision Tree, Logistic Regression, and Naïve Bayes classifiers. NLP was also used for string cleaning and to automatically generate URLs for resources such as YouTube, Wikipedia, and Coursera. Consequently, all generated resources undergo a domain-informed validation protocol based on industry and academic checklists/benchmarks, ensuring their relevance and quality. The evaluation was conducted on a sample of 500 labeled instances using ChatGPT and available assessment data. The results show a strong classification accuracy (F1-score ≥ 0.8 for a well-represented field) in skill-level classification and field-level predictions across the computing fields/subfields that indicate moderate accuracy (with a weighted F1 = 0.59), reflecting real-world data imbalance, and producing relevant resource sets that are topically consistent with the predicted computing specializations and career goals. By bridging predictive modeling with semantic web navigation, ComReF offers a suitable solution for computing field recommendations and helps reduce gaps in traditional advising approaches.

@article{Usman2026,
  title     = {AI-Driven Personalized Learning Framework for Computing Fields Recommendations in Secondary and Pre-University Contexts},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {17},
  number    = {9},
  year      = {2026},
  publisher = {The Science and Information Organization},
  author    = {Anas Inuwa Usman and Shahida Sulaiman and Ruhaidah Samsudin},
  doi       = {10.14569/IJACSA.2026.0170975},
  url       = {https://doi.org/10.14569/IJACSA.2026.0170975}
}