证券行业数据挖掘系统开发与设计

发布时间：2018-06-22 06:21

本文选题：证券业务 + 数据仓库　；参考：《电子科技大学》2013年硕士论文

【摘要】：证券行业中的各类数据是证券公司积累的宝贵财富，然而，很多证券公司并没有意识到这些数据的价值，静态的数据无法产生应有的效益，只有将不同类型的数据采用某种技术手段联系到一起，才能够发掘出数据之间内在的联系，更好地为证券公司的业务发展服务。利用数据仓库、联机分析处理、数据挖掘等方法，对证券公司的各类数据进行处理，目的是增强证券公司在客户关系管理和业务拓展决策方面的能力，并利用软件工程的相关理论和技术，开发了证券公司的业务管理分析系统。首先对证券公司在客户数据管理业务流程、行情与交易数据处理流程、业务营运数据处理流程等方面的需求进行了分析，明确了系统开发的目标，并对系统使用过程中的性能指标进行了研究。然后，，将系统划分为三层结构，分别是前台操作与结果展示模块、后台数据处理模块、数据源及数据仓库模块。前台模块的主要功能是实现系统与用户之间的人机交互功能，包括各类参数的设置、数据选择与处理、数据仓库构建操作、联机分析处理和数据挖掘结果的展现等；后台模块的主要是以服务的形式封装了数据联机分析处理服务、数据挖掘服务、ETL服务等功能，并根据系统的应用目标，实现了三类数据处理的业务逻辑。设计实现了系统的数据仓库，分别设计了数据仓库的概念模型、逻辑模型、物理模型。不同的数据处理与挖掘模块所应用的数据挖掘方法也不同，系统将根据实际的使用需求选择合适的数据联机分析和数据挖掘方法，利用决策树技术实现了客户关系管理数据的挖掘功能，利用概率神经网络技术实现了行情预测的数据挖掘功能。证券行业数据挖掘系统的构建，有利于实现对公司管理层决策的辅助支持，在实际的应用中能够提高风险的控制能力，争取客户源，实现公司利益的最大化。
[Abstract]:All kinds of data in the securities industry are valuable wealth accumulated by securities companies. However, many securities companies do not realize the value of these data, and static data can not produce the benefits they deserve. Only by connecting different types of data with some technical means can we find out the internal relationship between the data and better serve the business development of securities companies. By using the methods of data warehouse, on-line analytical processing, data mining and so on, this paper deals with all kinds of data of securities companies in order to enhance the ability of securities companies in customer relationship management and business expansion decision-making. Using the theory and technology of software engineering, the business management analysis system of securities company is developed. First of all, the requirements of securities companies in customer data management business process, market and transaction data processing process, business operation data processing process and so on are analyzed, and the goal of system development is clarified. The performance index of the system is studied. Then, the system is divided into three layers: foreground operation and result display module, background data processing module, data source and data warehouse module. The main function of the foreground module is to realize the man-machine interaction function between the system and the user, including the setting of all kinds of parameters, data selection and processing, data warehouse construction operation, on-line analysis processing and the display of data mining results. The background module mainly encapsulates the data on-line analytical processing service and the data mining service ETL service in the form of service, and realizes three kinds of business logic of data processing according to the application goal of the system. The concept model, logic model and physical model of the data warehouse are designed and implemented. Different data processing and mining modules have different data mining methods. The system will choose the appropriate data on-line analysis and data mining methods according to the actual needs. The decision tree technology is used to realize the function of customer relationship management data mining, and the probabilistic neural network technology is used to realize the data mining function of market forecast. The construction of securities industry data mining system is helpful to realize the auxiliary support to the company management decision, to improve the risk control ability in the practical application, to strive for the customer source, to realize the company benefit maximization.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP311.13;TP311.52

【相似文献】