基于用户行为分析的EMS代收货款业务优化

发布时间：2018-01-25 13:35

本文关键词： 代收货款 EMS 物流配送用户行为分析回归分析社区发现信用分析　出处：《河北科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来,代收货款业务已经成为EMS企业的主要业务增长点之一,代收货款业务的特殊性使其在传统的物流过程中附加了资金流。资金流的引入使得企业可以掌握大量的资金,但也加大了管理的难度。同时,各大物流企业代收货款业务的竞争也越发激烈,各企业也分别从人员配置、系统安全、物流配送、人性化服务等方面改进代收货款业务模式。本文以提升代收货款业务同城物流配送成功率为目标,主要做了如下三个工作:(1)分析了代收货款的产生背景、发展现状、主要的业务模式以及当前的机遇和挑战。确定以代收货款业务中的用户行为分析为出发点,改善其同城配送的服务质量。(2)从两个方面对用户行为进行了分析,其一是用户信用分析,通过对大量交易记录的整理,采用神经网络模型对代收货款过程中的三类主要的参与者(揽投员,发货方和收件方)建立信用评价体系;其二是用户趋势分析,针对发货方的发货量,通过关键点检测方法发现用户发货量的时序数据中的关键点,分析关键点的成因以加深对业务规律的理解,通过线性回归方法对用户未来一段时间的发货量进行一定程度的预测,为响应发货变化提供依据。(3)以信用评级和趋势预测为前提,本文构建了代收货款的用户网络,并在此网络上使用社区发现算法,挖掘针对揽投员的物流配送服务子网络,通过引入个性化的配送体系,改善了代收货款服务质量。结合具体案例,验证了该方法的有效性。
[Abstract]:In recent years, the collection of loans on behalf of the business has become one of the main business growth points of EMS enterprises. The particularity of the collection of payment for goods makes it attach capital flow to the traditional logistics process. The introduction of capital flow enables the enterprise to master a large amount of funds, but also increases the difficulty of management. At the same time. The competition between the major logistics enterprises to collect the payment for goods is becoming more and more fierce, and the enterprises are also from the personnel configuration, system security, logistics distribution. This paper aims at improving the success rate of logistics distribution in the same city, mainly doing the following three work: 1) analyzing the background of the generation of the collection of goods. Development status, the main business models and current opportunities and challenges. To determine the collection of loans in the business of user behavior analysis as the starting point. Improve the service quality of distribution in the same city. 2) analyze the user behavior from two aspects, one is the user credit analysis, through a large number of transaction records collation. The neural network model is used to establish a credit evaluation system for the three main participants in the process of receiving payment for goods (bid takers, shippers and recipients). The second is the user trend analysis, through the key point detection method to find the key points in the timing data of the user shipments, and analyze the causes of the key points in order to deepen the understanding of the business law. Through the linear regression method to predict the future shipments of users to a certain extent, to provide a basis for responding to changes in the delivery of goods, based on credit rating and trend prediction as the premise. In this paper, we construct the user network of collecting payment, and use community discovery algorithm on this network to mine the sub-network of logistics distribution service for the bid takers, and introduce the personalized distribution system. The quality of service is improved. The effectiveness of this method is verified by a case study.
【学位授予单位】：河北科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F259.2

【参考文献】