Analysis of Interprovincial Trade Relationships Based on Improved Gravity Modeling

Abstract

This paper analyzes the inter-provincial trade flows estimated by the improved gravity model using the complex network method. The article collects the distance and annual GDP value of each province from Baidu map and China Statistical Yearbook, and uses the gravity model to estimate the inter-provincial trade flow. Considering that the inter-provincial trade in the current year may be related to the GDP value of the previous year, the original model is improved by introducing the Gi-1 index and the results are normalized by multiplying them by 100 as the weights of the network constructed, and a fully connected network is established with the name of each province as a node connected network, and do the analysis of the network's node strength, weighted tight centrality, weight entropy of the three indicators, while using spectral clustering to further explore the trade relationship between the provinces.