Abstract

Abstract

Background:The growing burden of non-communicable diseases associated with poor nutrition and insufficient intake of essential nutrients, particularly in urban populations reliant on processed foods has caused a lot of ill health to the populace. Information is wealth as the saying goes. Many at times people may tend to forget the appropriate diets to take at any time. This could be as a result of their busy schedule or total lack of information on dietary ethics. This study presents the design and implementation of a persuasive mobile health application aimed at promoting the intake of dietary fiber and yeast-rich meals.Methodology: A mixed-methods approach was adopted, involving user surveys and system prototyping. The application incorporates lightweight convolutional neural networks (CNNs) for on-device meal recognition, enabling automatic logging of fiber and yeast content. Persuasive strategies such as SMS notifications, in-app alerts, and email reminders based on behavior-change theories were integrated to enhance user engagement and adherence. These reminders are designed to prompt users to log meals, reach daily targets, and stay informed about their nutrient intake. Results:The CNN achieved 92% accuracy in fiber classification and 89% in yeast classification. Functional testing confirmed the effectiveness of features such as automated logging, nutrient tracking, and adaptive recommendations. A pilot user survey indicated that 76% of participants became more aware of their intake, and 70% found the persuasive features motivating. Conclusion:This research demonstrates how machine learning and persuasive design can support healthier dietary habits and address specific nutrient deficiencies.

Keywords: Dietary, fibre, Mobile-health-application, persuasive, Saccharomyces cerevisiae, Yeast,