Emotion-Aware AI Meal Recommendation with Calorie-Constrained Recipe Scoring and Dual-Stage Deep Learning Models

Thonikkara, Mohammed Nihal (2025) Emotion-Aware AI Meal Recommendation with Calorie-Constrained Recipe Scoring and Dual-Stage Deep Learning Models. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (685kB) | Preview
Preview	PDF (Configuration Manual) Download (374kB) | Preview

Abstract

In fact, the contemporary world has grown used to the existence of smart solutions that can help justify everyday choices such as the food options that are determined due to the mood, the health objectives, and the convenience. This study project offers a simplified yet effective AI-based meal recommendation software that combines emotional insights and nutritional restrictions. The essence of the innovation is that it has a two-stage deep learning pipeline. The initial phase proposes a high-precision Bi-Directional Long ShortTerm Memory (Bi-LSTM) model to classify the emotion based on the text of the user with a test accuracy up to 92.87. This emotional inference is an exquisite feature engineering technique, which allows the system to encode affective cues in the downstream decision-making. The second phase introduces a lightweight regression-based model that produces a preference score of every recipe on combining the inferred emotive status with the energy needs. This model achieves a Mean Absolute Error (MAE) of 0.189, which proves that this model is effective in enabling the joint effects of emotional context and nutritional needs to determine the perceived suitability of recipes. In spite of the fact that the current system is more based on emotion-oriented personalisation and calorie matching, the methodology offers a basis of future extensions, such as ingredient replacement, deeper contextual perception and multimodal input. The obtained insights provide empirical support on the potential viability of emotionally aware meal recommendation and a solid foundation of the intelligent food-planning assistant of the next generation.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Razzaq, Abdul UNSPECIFIED
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence B Philosophy. Psychology. Religion > Psychology > Emotions Q Science > QP Physiology > Nutrition
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	02 Sep 2026 10:49
Last Modified:	02 Sep 2026 10:49
URI:	https://norma.ncirl.ie/id/eprint/9769

Actions (login required)

View Item