A Semantic Trajectory Mining System Based on Deep Neural Network
Xinyuan Zhang, Jiying Peng, Xinqi Zhang · 2021
At the beginning of 2020, epidemic of COrona VIrus Disease 19 (COVID-19) broke out. During the epidemic prevention and control, artificial intelligence, big data and other technologies have become powerful weapons against the epidemic, and have been widely used in the fields of epidemic tracing, confirming virus transmission path, resource allocation and so on. In this study, BiLSTM-CRF model, Bootstrap and Tornado frameworks are used to implement a neural network-based semantic trajectory mining system for the COVID-19 patients. On the basis of collecting the data published by the health committees of various provinces and cities, the semantic trajectories of the patients are extracted to ensure the accuracy of the data and then establish mapping relationship between the real space and the text description of the trajectories of the patients, while taking the time and space factors into account and excavating the dynamic changes of the patients.