Research On Monitoring And Early-Warning Of TBT Risk In Textile And Garment

Since China's entry into WTO, economic globalization and liberalization of trade have been on acceleration, and Technical Barriers to Trade (TBT) has become the main barrier in China's export. In the 1970s, TBT took up only 20% of non-tariff barrier, however, the percentage had risen to 45% in the 1990s and 80% up to now. With the influence of TBT from developed countries, our export products have been barricaded, export enterprises have been restricted, and China has suffered great economic loss. Furthermore, the situation is even worse in textile and garment industry.The main reason for the loss due to TBT is the ill-form of domestic enterprises, which makes it an urgent task to establish a early warming and alert system to collect and track foreign TBT dynamics. This system will enable enterprises to find potential TBT risk much earlier, so that the loss can be decreased to the least.In order to realize early warning and alert, the system must be supported by a speedy and effective information-processing mechanism. Deep research has been made with the TBT data released by Rapex that how to use automatic collection, data processing, data warehouse(DW), data mining(DM) in the early warning and alert system to solve the TBT problems. Automatic collection is based on the Train Collection Tool, and five steps including website selection, URL collection, content collection, saving settings, task scheduling are discussed. Though these steps, automatic collection process is well conducted, and structured data is output. The objective of automatic data processing is to transfer data collected into classified or digital information which can be identified and processed by computer. Data warehouse is introduced to help data organization though integrating historical data related to textile and garment TBT into the data warehouse. It is conducted in the way of data-driven and demand-driven. European Union TBT database is build based on the information acquired. And then, data warehouse is build basing on the measurements of number of recalls, growth rate of recalls per month, deviated growth rate of recalls per month, and on the dimensions of recalled date, sponsoring countries, objective countries, recalled products attribute, technical items, which is oriented by improving the competence of our textile and garment products. Data mining is used to explore hidden laws which are useful for early warning of TBT hazard. In this article, clustering algorithm and association algorithm are employed to find relations between level of risk and some of the influencing factors, so as to determine warning signal time, weight of each factor in the warning signal and model for early warning of TBT in textile and garment products. The model can help our government, association, enterprise to identify the potential risk and take corresponding measures ahead. At the end of the article, main achievements are summarized and subsequent work are suggested.