Intrinsic and Extrinsic Motivation Modeling Essential for Multi-Modal Health Recommender Systems
This addresses the problem of improving health outcomes through personalized guidance for individuals, but it appears incremental as it applies existing motivation concepts to the health domain.
The paper tackles the lack of effective health recommender systems by proposing an approach that models intrinsic and extrinsic motivations from multi-modal data to provide personalized lifestyle recommendations, aiming to enhance long-term engagement and sustainable behavioral adaptation.
Managing health lays the core foundation to enabling quality life experiences. Modern computer science research, and especially the field of recommender systems, has enhanced the quality of experiences in fields such as entertainment, shopping, and advertising; yet lags in the health domain. We are developing an approach to leverage multimedia for human health based on motivation modeling and recommendation of actions. Health is primarily a product of our everyday lifestyle actions, yet we have minimal health guidance on making everyday choices. Recommendations are the key to modern content consumption and decisions. Furthermore, long-term engagement with recommender systems is key for true effectiveness. Distinguishing intrinsic and extrinsic motivations from multi-modal data is key to provide recommendations that primarily fuel the intrinsic intentions, while using extrinsic motivation to further support intrinsic motivation. This understanding builds the foundation of sustainable behavioral adaptation for optimal personalized lifestyle health benefits.