GameMoodNet: Toward Adaptive Gaming via Multimodal Deep Affective Intelligence Framework

Abstract

In recent years, understanding player behavior and emotional states during gameplay has become an important research area in game analytics and human computer interaction. Emotional state detection enables game designers and researchers to analyse player engagement, improve user experience, and develop adaptive gaming environments. Traditionally, emotional behavior analysis in games relied on manual approaches such as questionnaires, player observation, and basic statistical analysis. These methods were time-consuming, less accurate, and unable to provide real-time insights into player emotions and behavior during gameplay. To address these limitations, this research proposes a Hybrid Fuzzy Boosting Architecture for Accurate Emotional State Detection in Real-Time Gameplay. The system is implemented as a web-based application using the Django web framework. Machine Learning (ML) techniques are used to analyse gameplay behavior data and predict emotional states efficiently. The research utilizes several ML algorithms including K-Nearest Neighbours (KNN), Random Forest (RF), Support Vector Machine (SVM), and a proposed hybrid model Fuzzy Neuro Boosting (FNB) model that combines Fuzzy Neural Network with Histogram Gradient Boosting (FNN-HGB). These algorithms are trained and evaluated using both classification and regression tree (CART) techniques to classify player behaviour (play_behavior) and predict engagement intensity (activity_level). Among the evaluated models, the proposed hybrid FNB model achieves higher prediction accuracy compared to traditional ML classifiers, demonstrating improved performance in handling complex and noisy gameplay data. The system also includes modules such as user authentication, exploratory data analysis, model training, performance comparison, and real-time prediction. By integrating ML models with a Django-based web interface, the developed system provides an efficient platform for analysing gameplay behaviour and predicting emotional states, thereby supporting better decision-making and improved player experience in modern gaming environments.