本文關鍵詞：銀行個人客戶挖掘與營銷系統(tǒng)　出處：《西安電子科技大學》2015年碩士論文　論文類型：學位論文

  更多相關文章： 數(shù)據(jù)倉庫 數(shù)據(jù)集市 數(shù)據(jù)倉庫建模 數(shù)據(jù)抽取/轉換/裝載 系統(tǒng)實現(xiàn)

【摘要】：個人客戶是銀行的主要客戶群體。隨著銀行業(yè)競爭的日益激烈,互聯(lián)網(wǎng)金融對銀行的沖擊越來越大,針對不同客戶特點,為客戶提供不同的服務和推薦客戶更感興趣的產(chǎn)品已成為銀行制勝的法寶之一。銀行通過多年聯(lián)機事務系統(tǒng)的運營積累了大量的數(shù)據(jù),缺少一套通過對數(shù)據(jù)進行分析和挖掘,幫助業(yè)務營銷的系統(tǒng)。為解決此問題,分行啟動了個人客戶挖掘與營銷系統(tǒng)的建設,該系統(tǒng)通過對個人客戶相關信息的整理、分析和挖掘,為業(yè)務營銷提供支持,同時為管理機構提供數(shù)據(jù)統(tǒng)計等功能和為業(yè)務決策提供幫助。本文首先討論了數(shù)據(jù)倉庫、數(shù)據(jù)集市和相關技術,分析了傳統(tǒng)數(shù)據(jù)庫和數(shù)據(jù)倉庫的差異,探討了數(shù)據(jù)倉庫的分類和系統(tǒng)結構。然后,從用戶的需求入手使用面向對象的分析技術通過類圖、用例模型、順序圖等方法,分析了系統(tǒng)實現(xiàn)需要完成的主要工作。分析時還從非功能性需求方面對系統(tǒng)的性能、數(shù)據(jù)安全、數(shù)據(jù)備份與恢復策略進行了考慮。對總行和分行的數(shù)據(jù)進行分析,明確了建立數(shù)據(jù)集市所需的數(shù)據(jù)源。其后,根據(jù)數(shù)據(jù)倉庫建模技術對系統(tǒng)的數(shù)據(jù)層進行了概念模型、邏輯模型和物理模型的三個層級的抽象和分析,最終設計了省分行從屬型數(shù)據(jù)集市的“兩級、三層”的系統(tǒng)結構,其中“兩層”是指數(shù)據(jù)的存儲的劃分,“三層”是指系統(tǒng)的實現(xiàn)結構劃分。隨后,根據(jù)系統(tǒng)分析的結果從數(shù)據(jù)抽取/轉換/裝載、數(shù)據(jù)挖掘、數(shù)據(jù)展現(xiàn)等方面對系統(tǒng)實現(xiàn)的關鍵模塊進行了設計。系統(tǒng)通過基于關聯(lián)規(guī)則挖掘實現(xiàn)產(chǎn)品關系的規(guī)則庫,當對某一具體客戶進行營銷時,系統(tǒng)根據(jù)客戶已持有的產(chǎn)品在規(guī)則庫中進行匹配,將匹配的結果推薦給客戶。通過數(shù)據(jù)挖掘的手段提高產(chǎn)品推介時的成功率,進一步提升用戶感受。最后,對系統(tǒng)實現(xiàn)過程中所使用的軟硬件環(huán)境進行了簡單的說明,展示了系統(tǒng)運行后的各個模塊的實際運行效果和在數(shù)據(jù)安全方面所做的工作。通過本次系統(tǒng)設計到實現(xiàn),初步探索了一套在國內銀行業(yè)有效的分行級數(shù)據(jù)集市建模方法。業(yè)務部門在對系統(tǒng)驗收后反饋:系統(tǒng)已基本滿足業(yè)務需求,完成了既定的目標。通過對系統(tǒng)的分析后續(xù)還可以進一步的對數(shù)據(jù)進行挖掘,同時增加更多維度的數(shù)據(jù)分析,讓數(shù)據(jù)在業(yè)務營銷中發(fā)揮更大的作用。本次系統(tǒng)的實現(xiàn)也為后續(xù)面向其他主題分析系統(tǒng)的建設積累了經(jīng)驗,為逐步完成企業(yè)級數(shù)據(jù)倉庫奠定了良好的基礎。
[Abstract]:Individual customers is the main customer groups of banks. The banking industry increasingly fierce competition, the Internet financial impact on banks is more and more big, for different customers, provide different services and recommend that customers are more interested in the products to customers has become one of the magic weapon for winning bank. The bank has accumulated a large amount of data through years of online transaction the lack of a system operation, through the analysis and the data mining system, to help business marketing. In order to solve this problem, the branch started the construction of individual customer mining and marketing system, the system through the relevant information of the individual customer collation, analysis and mining, provide support for business marketing, data statistics and other functions for the management of institutions and provide help for business decisions at the same time. This paper discusses data warehouse, data mart and related technology, the traditional database and data analysis Warehouse differences, discusses the classification and system structure of data warehouse. Then, from the user's demand of using the technology of object-oriented analysis by class diagram, use case model, sequence diagram analysis method, the system needs to complete the work. The analysis of system performance, from the non functional requirements of data security, data backup and recovery strategies were considered. The head office and the branch of the data analysis, clearly needed to establish the source of data for the data marts. Subsequently, according to the data of data warehouse modeling technology on the system layer of the conceptual model, three levels of abstraction and analysis of logical model and physical model, the final design the branch of the subordinate data mart "two levels, three layer system structure", the "two layer" refers to the division of data storage, "three" refers to the realization of the system structure is classified. Then, root According to the results from the analysis of the data extraction system of ETL, data mining, data display and other aspects of the design of the key modules in this system. The system through the rule base to achieve product relationship based on association rule mining, when marketing to a specific customer, the customer has to hold the system according to the matching rules in products in the library, the matching results to recommend to customers. By means of data mining to improve the success rate of product promotion, to further enhance the user experience. Finally, the system hardware and software environment in the process of using a simple description, showing the effect of actual operation of each module of the system after running and done in data security. This system is designed to achieve, explore a set of effective in the domestic banking branch level data mart modeling method. Business Department of the Department of The acceptance of feedback: the system can satisfy the needs of the business, to complete the established goals. Through the analysis of the system can also carry out further follow-up of data mining, while increasing the analysis more dimensions of data, make data play a more important role in business marketing. Implementation of this system for the follow-up for other topics analysis of accumulated experience of system construction, and laid a good foundation for the gradual completion of enterprise data warehouse.

【學位授予單位】：西安電子科技大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TP311.13

【參考文獻】

相關期刊論文 前8條

1 姚敏;萬里勇;;高級數(shù)據(jù)挖掘與應用[J];國際學術動態(tài);2015年01期

2 汪樟發(fā);涂俊;張春霞;;大數(shù)據(jù)商業(yè)模式與傳統(tǒng)企業(yè)的轉型創(chuàng)新[J];創(chuàng)新與創(chuàng)業(yè)管理;2014年02期

3 曹軍;;淺析數(shù)據(jù)倉庫在ERP系統(tǒng)中的應用[J];青春歲月;2013年14期

4 蔣勇青;楊奕虹;楊賀;;論數(shù)據(jù)清洗對信息檢索質量的影響及清洗方法[J];中國索引;2012年01期

5 劉頎桓;;基于數(shù)據(jù)倉庫的商業(yè)銀行報表系統(tǒng)基礎數(shù)據(jù)分析設計[J];企業(yè)技術開發(fā);2011年23期

6 王磊;;數(shù)據(jù)挖掘技術在電子商務中的應用[J];青春歲月;2011年08期

7 孟娟;;數(shù)據(jù)挖掘在銀行業(yè)中的應用[J];大眾商務;2010年02期

8 李小慶;;銀行數(shù)據(jù)挖掘系統(tǒng)的構建思路[J];中國金融電腦;2008年08期

，

本文編號：1369326