Discovering Environmental Impacts on Public Health Using Heterogeneous Big Sensory Data

In this paper, we present a method for detecting events, especially healthcare-related events, by abstracting trends of data streaming from heterogeneous sensors. The main idea behind the method is to detect real-time events and explain them understandably by finding spatio-temporal-theme correlations between physical and social sensory data. In the method, a training stage is designed as a non-stop process with labels assigned automatically to feature vectors in order to build a set of positive and negative samples. Thereafter, an event model is generated by using supervised learning approaches as a means to steadily increase its accuracy. The problem of environmental impacts on asthma attacks is used to evaluate the proposed method. Experimental results show that the proposed method can detect the prevalence of asthma risks in a specific spatio-temporal context with high accuracy.