传染病监测数据中高危区域和潜在传染邻域探测.docVIP

 Detection of High-risk Zones and Potential Infected Neighbors from Infectious Disease Monitoring Data# 5 10 15 20 25 30 35 40 45 TAN Biying1, DUAN Lei1, GOU Chi1, HUANG Shuyang1, FANG Yuhao1, ZHAO Xing2, TANG Changjie1** (1. School of Computer Science, Sichuan University, Chengdu 610065; 2. West China School of Public Health, Sichuan University, Chengdu 610041) Abstract: Detecting the high-risk zones as well as potential infected geographical neighbor is necessary and important to reduce the loss caused by infectious disease. However, it is a challenging work, since the outbreak of infectious disease is uncertain and unclear. Moreover, the detection should be efficient otherwise the best control and prevention time may be missed. To deal with this problem, we propose a geography high-risk zones detection method by capturing the significant change in the infectious disease monitoring data. The main contribution of this paper includes: (1) Analyzing the challenges of the early warning and detection of infectious disease outbreak; (2) Proposing a method to detect the zone that the number of monitoring cases changes significantly; (3) Defining the infection perturbation to describe the infection probability between two zones; (4) Designing an algorithm to measure the infection perturbation of infectious disease between adjacent zones; (5) Performing extensive experiments on both real-world data and synthetic data to demonstrate the effectiveness and efficiency of the proposed methods. Key words: Data Mining; Change Mining; Spatial Mining; Time Series 0 Introduction The issue of infectious disease prevention and control is important in public health and humans’ daily life. For example, the acute piratory syndrome (SARS) outbreak in 2002 caused huge losses. According to the report from WHO, 8069 people infected SARS and 775 people lost lives in 32 main countries of the world since its outbreak to July.11th ,2003. Not only the direct loss but also the indirect loss