【摘要】：隨著經(jīng)營(yíng)管理水平的不斷提升,黑龍江省農(nóng)村信用社在客戶管理、風(fēng)險(xiǎn)管理、運(yùn)營(yíng)管理、財(cái)務(wù)管理、監(jiān)管合規(guī)等多個(gè)業(yè)務(wù)領(lǐng)域的需求越來越多,對(duì)數(shù)據(jù)的使用提出了更高的要求。黑龍江省農(nóng)村信用社經(jīng)過多年的系統(tǒng)建設(shè),積累了大量的業(yè)務(wù)數(shù)據(jù),然而,由于早期農(nóng)信社的各系統(tǒng)的建設(shè)是相對(duì)獨(dú)立的,缺乏完善的數(shù)據(jù)管控體系以及對(duì)數(shù)據(jù)的統(tǒng)一管理,導(dǎo)致了農(nóng)信社在數(shù)據(jù)存儲(chǔ)中出現(xiàn)了沒有統(tǒng)一的數(shù)據(jù)源、業(yè)務(wù)數(shù)據(jù)分散、各業(yè)務(wù)系統(tǒng)數(shù)據(jù)不一致、數(shù)據(jù)質(zhì)量不高、安全性低等問題。現(xiàn)有的分散的數(shù)據(jù)不能滿足企業(yè)級(jí)精細(xì)化管理和個(gè)性化營(yíng)銷的需求。因此需要建立數(shù)據(jù)平臺(tái),來挖掘并發(fā)揮企業(yè)數(shù)據(jù)的最大價(jià)值。本課題的研究正是著眼于目前農(nóng)信社數(shù)據(jù)存儲(chǔ)的弊端,著重研究了數(shù)據(jù)平臺(tái)系統(tǒng)在農(nóng)信社系統(tǒng)中的實(shí)際應(yīng)用。本文以黑龍江省農(nóng)村信用社數(shù)據(jù)平臺(tái)項(xiàng)目為實(shí)際項(xiàng)目背景,主要闡述了數(shù)據(jù)平臺(tái)項(xiàng)目一期基礎(chǔ)平臺(tái)的設(shè)計(jì)與實(shí)現(xiàn)過程。通過對(duì)農(nóng)信社歷史數(shù)據(jù)及實(shí)際業(yè)務(wù)經(jīng)營(yíng)情況的分析得出了黑龍江省農(nóng)村信用社數(shù)據(jù)平臺(tái)數(shù)據(jù)存儲(chǔ)模型,整理并制定了統(tǒng)一的業(yè)務(wù)數(shù)據(jù)標(biāo)準(zhǔn)格式,并采用OLAP(On-Line Analytical Processing,聯(lián)機(jī)在線分析處理)、透明網(wǎng)關(guān)等技術(shù)設(shè)計(jì)并完成了所有業(yè)務(wù)數(shù)據(jù)自源系統(tǒng)生成至存儲(chǔ)到數(shù)據(jù)模型過程中所涉及的數(shù)據(jù)采集、數(shù)據(jù)集成加工等功能模塊,以及業(yè)務(wù)數(shù)據(jù)自數(shù)據(jù)模型中提取出來到展現(xiàn)在最終用戶的過程中涉及的數(shù)據(jù)聚合,數(shù)據(jù)展現(xiàn)及分發(fā)等功能模塊,并且針對(duì)數(shù)據(jù)平臺(tái)內(nèi)全部業(yè)務(wù)數(shù)據(jù)流轉(zhuǎn)的E(Extract)、T(Transformation)、L(Load)過程各個(gè)環(huán)節(jié)的實(shí)際情況,采用了有針對(duì)性的處理方式完成了ETL調(diào)度模塊的設(shè)計(jì)與實(shí)現(xiàn)。通過詳細(xì)測(cè)試本平臺(tái)上游各業(yè)務(wù)系統(tǒng)的數(shù)據(jù)能夠成功的加載到數(shù)據(jù)平臺(tái)系統(tǒng)中,而且進(jìn)入數(shù)據(jù)平臺(tái)之后經(jīng)過數(shù)據(jù)加工處理,數(shù)據(jù)平臺(tái)中各層之間的數(shù)據(jù)比對(duì)正確;系統(tǒng)的調(diào)度平臺(tái)能夠按照各工作流的預(yù)先配置自動(dòng)發(fā)起數(shù)據(jù)抽取、轉(zhuǎn)換、加載等任務(wù),對(duì)業(yè)務(wù)數(shù)據(jù)進(jìn)行批量處理,并且能夠完成對(duì)任務(wù)處理的監(jiān)控與統(tǒng)計(jì),很好的完成數(shù)據(jù)平臺(tái)項(xiàng)目的需求。
[Abstract]:With the continuous improvement of management level, the demand of Heilongjiang rural credit cooperatives in many business areas, such as customer management, risk management, operation management, financial management, regulatory compliance, and so on, has put forward higher requirements for the use of data. The rural credit cooperatives in Heilongjiang province have accumulated a large amount of business after years of system construction. Data, however, due to the relatively independent construction of the early agricultural credit cooperatives, the lack of perfect data management system and the unified management of data, the agricultural credit cooperatives in the data storage have no unified data sources, business data dispersion, different business system data, data quality is not high, low security and so on. The existing scattered data can not meet the needs of the fine enterprise management and personalized marketing. Therefore, it is necessary to establish a data platform to excavate and exert the maximum value of the enterprise data. The research of this subject is focusing on the disadvantages of the data storage of the agricultural credit cooperatives at present, and focuses on the research of the reality of the data platform system in the agricultural credit cooperative system. Using the data platform project of Heilongjiang rural credit cooperatives as the actual project background, this paper mainly expounds the design and implementation process of the basic platform of the data platform project. Through the analysis of the historical data and the actual business operation of the rural credit cooperatives, the data storage model of the data platform of the rural credit cooperatives in Heilongjiang is obtained. The unified business data standard format is formulated, and the OLAP (On-Line Analytical Processing, online online analysis and processing) and transparent gateway are used to design and complete the data collection, data integration processing and other functional modules, as well as the number of services involved in the process of generating all the business data from the source system to the data model. According to the self data model, the function modules such as data aggregation, data presentation and distribution involved in the process of end-user are extracted, and the actual situation of each ring of the E (Extract), T (Transformation), L (Load) process in the data platform is carried out by a targeted processing method. The design and implementation of the L scheduling module can be successfully loaded into the data platform system by testing the data of the upstream business system in this platform, and after data processing, the data alignment between each layer in the data platform is correct; the system scheduling platform can be configured in advance according to the workflow. It automatically initiates data extraction, conversion, loading and other tasks, batch processing of business data, and can complete the monitoring and statistics of task processing, and complete the requirements of the data platform project.
【學(xué)位授予單位】：哈爾濱工業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TP311.13

【參考文獻(xiàn)】

相關(guān)期刊論文 前2條

1 鐘華;馮文瀾;譚紅星;黃濤;;面向數(shù)據(jù)集成的ETL系統(tǒng)設(shè)計(jì)與實(shí)現(xiàn)[J];計(jì)算機(jī)科學(xué);2004年09期

2 胡振華;吳袁萍;;我國(guó)村鎮(zhèn)銀行現(xiàn)狀與發(fā)展探究[J];中國(guó)農(nóng)村科技;2011年07期

，

本文編號(hào)：2158869