Spatial Data Mining Assisting Urban Epidemic Surveillance with the Weighted DBSCAN Algorithm

Urban epidemic monitoring will become a normal work for a long time in the future. How to develop and utilize spatial data has become an important direction for studying the characteristics of epidemic transmission and predicting and controlling the development of epidemic. In view of the large volume, multi-source and multi-dimensional characteristics of location big data, the data mining method that comprehensively dealed with the geographical and non-geographical attributes of location data in the way of weight coefficient can more accurately discover high-traffic commercial network clusters in cities and avoid the omission of high-traffic commercial network. It provided a scientific basis for decision making to accurately discover the key areas with high transmission and to obtain the spatial distribution of the key areas with high transmission and infection. The simulation results showed that the weighted clustering algorithm proposed in this paper showed obvious optimization effect, which was in good agreement with the urban epidemic statistics released by the government, and played a certain role in supporting the effective development of the urban epidemic surveillance.