Fitness-based Recommender Systems for Reducing Sedentary Behaviour

Physical inactivity among the global population has warranted substantial concern in recent times. In 2020, the World Health Organization (WHO) announced that approximately 28% of adults and 81% of children were living sedentary lifestyles which could lead to chronic obesity. A wealth of literature demonstrates that long-term obesity increases risk factors related to increased mortality via cardiovascular, pulmonary, and cognitive diseases / decline.

Existing interventions such as mobile apps and fitness centers have been shown to be ineffective due to a one-size fits-all approach and ad hoc design principles. Furthermore, there are barriers related to location, accessibility, and a high cost of entry.

My research proposes to use classical machine learning, generative artificial intelligence (LLMs), and explainability to build trust-worthy tools that can generate persuasive, tailored fitness plans in real-time.