Development of an Intelligent Households Balanced Diet Advisory System for Women and Children (Vulnerable) in the North Eastern Region of Nigeria.
Publication Date: 11/11/2025
Author(s): Jeremiah Yusuf Bassi, Samuel Awuna Kile.
Volume/Issue: Volume 8, Issue 2 (2025)
Page No: 185-204
Journal: British Journal of Computer, Networking and Information Technology (BJCNIT)
Abstract:
According to UN estimates, approximately 4.4 million children in Nigeria's northeastern region suffer from acute malnutrition, a problem made worse by conflict and displacement. Even though the area is agrarian, with more than 83% of households growing crops, there is a big disconnect between the availability of food and nutritional awareness, which results in diets that are insufficient. By developing a smart family balanced food advisory system for vulnerable women and children, this study closes this gap. The research methodology used in the study is empirical and data-driven. In order to achieve nutrient requirements and minimize costs, a balanced diet model using a linear-program (LP) formulation was applied. To categorize dietary adequacy, the system uses a Support Vector Machine (SVM) model that was trained on regional dietary and socioeconomic data. With a high recall of 84.4% for the "inadequate" class and a 91.0% accuracy and a robust ROC-AUC of 0.963 in identifying homes at nutritional risk, the model showed great efficacy. Dietary diversity, energy, and protein intake were validated as important predictors by feature significance analysis. This work offers a scalable method to directly battle malnutrition by empowering communities with data-informed nutritional advise through the use of an AI-driven, scientifically validated tool that converts local food production into practical, balanced diet recommendations.
Keywords:
Advisory system, balanced diet, intelligent, machine learning, and support vector machines.
