数据挖掘在证券业务客户服务中的应用

发布时间：2018-02-12 04:42

本文关键词： 证券数据挖掘朴素贝叶斯分类客户服务　出处：《华侨大学》2017年硕士论文　论文类型：学位论文

【摘要】：在证券行业高速发展的今天,市场的竞争也相当激烈,原有靠收取佣金为主的传统模式已不再适应当今时代。提升客户服务质量,加强服务质量管控,提升客户感知、延长客户生命周期,转型为综合性金融服务公司已经成为各大证券公司在激烈竞争中获得优势的努力方向。本文结合华福证券泉州分公司的经营情况及数据,挑选了朴素贝叶斯算法进行数据挖掘研究。从客户海量的相关数据中提取资产、交易等数据,通过数据清洗归纳汇总、文本型数据数值化、连续数值离散化的方法,对数据进行预处理。根据业务需求及朴素贝叶斯算法的特点,最终选取资产、换手率、盈亏率这三个特征向量来进行数据挖掘,将客户分类为高价值客户、高价值易消亡客户、普通客户、低价值客户、无价值客户五类。在确定模型后,对模型分类结果进行验证及优化,准确率达到90%以上。通过本课题的研究,能准确地对客户进行分类,并根据业务特点,对分类出的不同类型的客户提供不同的服务,提高了客户服务的深度和准确性,为前台服务人员提供了强大的数据支持,有效提高了客户的保有能力,进而提升证券公司的竞争力,提高证券公司收入。通过各种数据挖掘技术的引入,比如本文的贝叶斯分类模型等,可以建立起高效的证券行业客户服务体系,并且完善全客户生命周期客户机制。针对客户特点引导客户投资,推动客户转型,最终促进证券公司转型。数据挖掘技术的引入将使证券公司客户服务工作更好更有效率。
[Abstract]:With the rapid development of the securities industry today, the competition in the market is also very fierce. The traditional mode of collecting commission is no longer adapted to the times. To improve the quality of customer service, to strengthen the control of service quality, to enhance customer perception, Extending the customer life cycle and transforming into a comprehensive financial service company has become the direction of each major securities company to gain an advantage in the fierce competition. This paper combines the management situation and data of Quanzhou Branch of Huafu Securities. This paper selects naive Bayes algorithm to study data mining. It extracts assets, transactions and other data from the massive data of customers, summarizes the data through data cleaning, numerical method of text data, and discrete method of continuous numerical value. According to the business requirements and the characteristics of naive Bayes algorithm, we select the three characteristic vectors of assets, turnover ratio and profit / loss ratio to mine the data, and classify the customers as high-value customers and high-value perishing customers. Five categories of ordinary customers, low value customers and no value customers. After determining the model, the model classification results are verified and optimized, and the accuracy is over 90%. Through the research of this topic, we can classify the customers accurately. And according to the characteristics of the business, it provides different services to different types of customers, improves the depth and accuracy of customer service, provides strong data support for the front desk service personnel, and effectively improves the customer retention ability. Through the introduction of various data mining technologies, such as the Bayesian classification model of this paper, we can establish an efficient customer service system in the securities industry. And perfect the customer mechanism of the whole customer life cycle, guide the customer investment according to the customer characteristics, promote the customer transformation, and finally promote the transformation of the securities company. The introduction of data mining technology will make the customer service work of the securities company better and more efficient.
【学位授予单位】：华侨大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F832.39;TP311.13

【相似文献】