A Survey on Ambient Intelligence Contexts: A Context-Aware Taxonomy based on Deep Learning and Internet of Things Synergy
Parsa Pure Hamedany · 2024
Deep Learning (DL) and the Internet of Things (IoT) are critical components of modern Ambient Intelligence (AmI), which integrate state-of-the-art Artificial Intelligence (AI) and Information Communication Technologies. The recent synchronization of IoT-enabled big data generation (5G) and advanced reasoning of DL models, such as Multimodal Large Language Models, has created new opportunities and challenges for modern AmI. However, comprehensively analyzing and understanding the broader implications of AmI in the big picture is difficult because of its interdisciplinary nature. To address this intricacy, by adopting the Context-Aware Computing perspective, a systematic arrangement of AmI contexts is proposed. This survey develops a taxonomy of AmI contexts based on four key dimensions: Human, System, Space, and Time. Each of these dimensions is further split into sub-context categories. By organizing DL and IoT applications within this taxonomy, the study offers a systematic framework to understand and customize AmI systems based on requirements. The resulting context portfolios serve as a flexible conceptual-functional toolkit for researchers and practitioners, aiding them in selecting and adapting contexts for specific applications. This taxonomy aims to clarify the complex landscape of AmI and provide a foundation for future innovations in the field.