基于索引结构的Apriori算法及其在外贸CRM中的研究应用

发布时间：2018-04-02 09:17

本文选题：客户关系管理　切入点：索引结构　出处：《哈尔滨商业大学》2013年硕士论文

【摘要】：从我国加入WTO以来,我国的经济实现了与国际接轨,市场经济竞争越来越激烈,企业要想生存乃至取得长远发展就需要先进的管理方法为其保驾护航。特别是外贸企业,其客户都是国外的企业,距离比较远,与客户联系大多不是面对面的交流,并且商品需要长距离运输,有繁琐的手续和成本支出。因此,确切的掌握客户的动向、了解客户的需求,不仅可以在一定程度上降低经济成本支出减少交易风险,还可以维系老客户、发展新客户。数据挖掘能够从海量的数据中提取出来人们感兴趣的信息,把企业的商品交易数据作为数据挖掘的对象就可以挖掘出以客户为导向的商品交易趋势,从而可以准确的把握客户的需求,使得企业的客户关系管理能够更加完善。在数据挖掘原有研究背景的基础上,深入分析关联规则的经典算法-Apriori算法,研究算法的基本思想,算法的性质,并对算法进行详细的性能分析。针对Apriori算法需要多次扫描数据库及需要储存大量候选项集等弊端,提出基于索引结构的Apriori改进算法。改进算法采用的存储方式与原有的Apriori算法不同,不是存储各个事务记录包含的候选项目,而是存储候选项目所对应的事务标识列表。通过对候选项目所对应的事务标识列表中的事务计数得到该候选项目的支持度。地址索引结构使得频繁项集的连接和事务标识列表之间的连接更加高效快速,从而提升了剪枝速度。论文详尽描述改进后的算法并进行实例分析,将原有算法和改进算法进行对比,通过在Eclipse平台上进行实验证明改进算法确实具有更高的时间效率。最后,将此改进算法应用于外贸企业客户关系管理系统中,以java语言为基础,在Eclipse平台上仿真实现了外贸企业客户关系管理系统。通过加入数据挖掘模块完善了普通外贸企业客户关系管理系统的功能,提高了此系统的智能性,使企业的管理决策水平与世界接轨。
[Abstract]:Since China's entry into WTO, the economy of our country has been in line with international standards, and the competition in the market economy has become more and more intense. If enterprises want to survive or even achieve long-term development, they need advanced management methods to protect them, especially foreign trade enterprises. Its customers are foreign enterprises, far away, contact with customers is not mostly face-to-face communication, and the goods need long distance transportation, there are cumbersome procedures and costs. Understanding the needs of customers can not only reduce economic costs, reduce transaction risks, but also maintain old customers and develop new customers. Data mining can extract information of interest from vast amounts of data. Taking the commodity trading data of the enterprise as the object of data mining, we can mine out the trend of the commodity trading oriented by customers, thus we can accurately grasp the needs of the customers, and make the customer relationship management of enterprises more perfect. On the basis of the original research background of data mining, this paper deeply analyzes the classical algorithm of association rules-Apriori algorithm, studies the basic idea of the algorithm and the nature of the algorithm. The Apriori algorithm needs to scan the database many times and store a large number of candidate items. This paper proposes an improved Apriori algorithm based on index structure, which uses a different storage method than the original Apriori algorithm, and is not a candidate for storing each transaction record. Instead, the list of transaction identifiers corresponding to the candidate items is stored. By counting the transactions in the list of transaction identifiers corresponding to the candidate items, the support degree of the candidate items is obtained. The address index structure enables the join of frequent itemsets and. Connections between transaction identification lists are more efficient and faster, In this paper, the improved algorithm is described in detail and an example is analyzed, and the original algorithm is compared with the improved algorithm. The experiment on Eclipse platform shows that the improved algorithm does have higher time efficiency. Finally, the improved algorithm is applied to the customer relationship management system of foreign trade enterprises, based on the java language. The customer relationship management system of foreign trade enterprise is simulated on Eclipse platform, and the function of customer relationship management system of common foreign trade enterprise is improved by adding data mining module, and the intelligence of this system is improved. Make the management decision level of the enterprise and the world in line with the world.
【学位授予单位】：哈尔滨商业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP311.13

【相似文献】