"Context-Aware Smart Home IoT Dataset for Energy and Behavioral Analytics"

"This work presents a large-scale synthetic smart home Internet of Things (IoT) dataset designed to support research in energy analytics, behavioral modeling, and intelligent system development. The dataset captures multivariate time-series data representing a context-aware smart home environment, including environmental variables such as temperature, light intensity, humidity, and gas concentration, along with behavioral features such as occupancy, motion, and visitor presence. Spatial variability is incorporated through multi-room sensing, while a physics-inspired energy model reflects realistic consumption patterns influenced by environmental conditions and human activity. Additionally, the dataset includes sensor imperfections such as noise, drift, missing values, and fault events to emulate real-world deployment scenarios. The dataset is provided in three variants\u2014complete, noise-free, and behavior-free\u2014to support controlled experimentation and ablation studies. With a total size of approximately one gigabyte and a scalable generation framework, the dataset enables both small-scale prototyping and large-scale machine learning applications. This dataset serves as a flexible and reproducible resource for smart home research, including occupancy prediction, energy forecasting, anomaly detection, and IoT system benchmarking."