The first is activity analysis. Every quarter, Baidu traffic statistics and other tools are used to monitor the visits, click-through rates, update frequency, number of users, etc. of e-commerce platform websites related to online shopping bonded import business in real time to analyze the activity of cross-border e-commerce platforms. The activity analysis results are compared with the online shopping bonded import business data to screen out high-risk e-commerce platform companies.
The second is concentration analysis. Relying on the big data analysis platform built by the functional departments, the orders, logistics orders, payment orders, declaration lists and other data of the entire chain of online shopping bonded import business are integrated, and the concentration analysis of the orderer, payer, orderer’s phone number, delivery area, sales category and other data is carried out. Highly concentrated and abnormal data such as orderers, payers, orderers’ phones are screened to form a batch of risk clues for subsequent disposal.
The third is credibility analysis. In accordance with the prosecution requirements of the “three-order comparison” of cross-border e-commerce, the data such as “same number (ID number) but different name of payer”, “same number (ID number) but different name of orderer”, “same number (phone number) but different person of recipient”, and “same person but different location of recipient” in the order, payment order, and logistics order are compared and analyzed to determine the credibility of the data reported by the enterprise. The enterprises and platforms with abnormal analysis results are listed as those with poor credibility, providing support for risk assessment and strengthening of actual supervision.
The fourth is correlation analysis. Based on massive lists, orders, logistics orders, payment information and other data, the correlation relationships of different business roles (warehousing within the district, e-commerce platforms, e-commerce enterprises, payments, guarantees, logistics and express delivery, etc.) scattered among them are displayed through group statistics. The illegal clues discovered by risk prevention and control and confirmed by subsequent audits can be screened out through correlation analysis to identify business-related enterprises with suspected illegal and irregular enterprises, and “follow the clues” to lock in and investigate other related enterprises.