Application Research Of Apriori Algorithm Based On Index Structure In CRM Of Foreign Trade

Since join in the WTO, the economy of our country has realized integration with global economy. The market economy competition is more intense, the survival and long-term development of enterprises need advanced management methods. Especially foreign trade enterprise, its customers are all foreign companies. The far distance makes communicate with customers face to face is very difficult. And the goods of foreign trade need long distance transportation and have a tedious procedure and costs. So accurately grasp customer’s tendency and understand the demands of customers not only can save economic costs and reduce transaction risk in a certain extent, but also can keep the old customers and develop new customers. Because of data mining can get information that people interested from mass data, we can get customer oriented trend of commodities trading if use commodity transaction data as mining object. Thus companies can accurately grasp the customer demands, and then makes its customer relationship management more perfect.Based on the original research of data mining, deeply analyze the classical association rules algorithm-Apriori algorithm, learn its basic ideas and nature, and then analyze its performance in detail. Because of Apriori algorithm needs scan database many times and store a large number of candidate itemsets, an improved Apriori algorithm based on index structure is put forward. The storage mean of improved algorithm is different from the original Apriori algorithm. It not stores the candidate projects that included in transaction records, but the lists of transaction id that candidate projects corresponding. Through counting the transactions in transaction id list that candidate projects corresponding can get the support of the candidate projects. Address index structure makes the connection of frequent itemsets and transaction id lists more efficient, and this promotes the speed of pruning. After describe and instance analysis the improved algorithm, comparing the original algorithm with the improved algorithm. Then proving the improved algorithm has higher time efficiency through doing experiments on Eclipse platform.Finally, applying this improved algorithm to foreign trade enterprise customer relationship management system, and then using java language, simulation realized the foreign trade enterprise customer relationship management system on Eclipse platform. Though adding data mining module perfect the functions of common customer relationship management system on foreign trade enterprises, and improved the intelligent of this system. All these make the level of enterprise in management decision-making is geared to international standards.