The invention relates to the technical field of urban data operation, in particular to a Shanghai address
database dynamic updating operation method, which comprises the following steps of: uniformly
processing data formats by completing docking with
data source interfaces of Shanghai
public security,
postal service and maps; existing address data in Shanghai are disassembled into hierarchical fields of cities, districts, streets, roads and roadways according to a 18-level
structured model, hierarchical differences between a new
data source and a golden address
library are continuously compared, address change authenticity is inferred through a
knowledge graph and a graph neural network, target addresses which are not inquired in continuous 3 months are judged as suspected discarding, and discarding of the target addresses is judged. And predicting a discarding probability and a change state in combination with a
time sequence model, and archiving historical data to obtain a complete address
library containing a life cycle. According to the method, the problems that the efficiency is low and intelligent, efficient and traceable address
library dynamic updating operation cannot be realized when a traditional regular full-quantity synchronization mode is used for dealing with high-frequency address change of an ultra-
large city in Shanghai are solved.