【摘要】：近幾年隨著物流成本的降低,網(wǎng)上購物的人數(shù)迅速增加,電子商務行業(yè)快速發(fā)展。人們開始重視持續(xù)性發(fā)展,對決策的制定有了更高的要求,要求有更合理有效的數(shù)據(jù)支持,這就需要對大量的銷售數(shù)據(jù)進行分析。數(shù)據(jù)倉庫就是在這樣的大環(huán)境下出現(xiàn)并且得到快速發(fā)展。使用數(shù)據(jù)倉庫,如何對數(shù)據(jù)分析,分析到如何程度的BI(Business Intelligence)系統(tǒng),對企業(yè)的作用至關(guān)重要。BI系統(tǒng)可以幫助企業(yè)建立一個清晰、準確的決策支持系統(tǒng),可以實現(xiàn)真正意義上的商務智能,而且利用存儲的公司日常運營所產(chǎn)生的銷售數(shù)據(jù)、采購數(shù)據(jù)、倉庫庫存等內(nèi)容,幫助公司實現(xiàn)更有效的資源配備和渠道管理。本系統(tǒng)首先使用ETL工具Kettle從各種數(shù)據(jù)源(如多個服務器上的MySQL數(shù)據(jù)庫、SQL Server數(shù)據(jù)庫、各種excel表格)抽取數(shù)據(jù)聚合到一個Oracle數(shù)據(jù)庫里,然后在Oracle中將數(shù)據(jù)按照粒度層級不同整合到CDC層、ODS層和DW層。在這三層之間通過調(diào)動存儲過程來轉(zhuǎn)化,對于不同的需求定制不同的存儲過程定時調(diào)用,將數(shù)據(jù)寫到結(jié)果表中,供前端數(shù)據(jù)展示模塊調(diào)用。平臺展示模塊使用兼容性較好的B/S架構(gòu),網(wǎng)站使用PHP語言進行開發(fā),頁面數(shù)據(jù)展示使用ExtJS。前端頁面能夠動態(tài)變化,以多種樣式展示數(shù)據(jù),包括折線圖、餅狀圖和柱狀圖等等,并且能夠?qū)δ骋惶囟〝?shù)據(jù)在不同時間的數(shù)值進行對比展示。經(jīng)過嚴格的測試表明,該系統(tǒng)提供的數(shù)據(jù)倉庫設計合理,能夠讓業(yè)務和管理人員和方便的獲取想要的數(shù)據(jù),可以極大地提高公司管理層的工作效率,具有較好的用戶體驗,為企業(yè)提供了較好的決策支持,具有良好的使用價值。
[Abstract]:In recent years, with the decrease of logistics cost, the number of online shopping has increased rapidly, and the e-commerce industry has developed rapidly. People begin to pay more attention to sustainable development, have higher requirements for decision-making, and demand more reasonable and effective data support, which requires the analysis of a large number of sales data. Data warehouse is in this kind of big environment appearance and get rapid development. Using the data warehouse, how to analyze the data, how to analyze the degree of BI (Business Intelligence) system is very important to the enterprise. BI system can help the enterprise to establish a clear and accurate decision support system. It can realize real business intelligence, and make use of the sales data, purchasing data, warehouse inventory and so on, which can help the company achieve more effective resource allocation and channel management. This system first uses the ETL tool Kettle to extract data from various data sources (such as MySQL database, SQL Server database on multiple servers, various excel tables) and aggregates the data into a Oracle database. Then the data is integrated into CDC layer, ODS layer and DW layer according to granularity level in Oracle. Through transferring the stored procedure between the three layers, different stored procedure timing calls are customized for different requirements, and the data is written to the result table for the front-end data display module to call. The platform display module uses the compatible B / S architecture, the website uses PHP language to develop, and the page data display uses ExtJS.. The front page can change dynamically, displaying data in a variety of styles, including line, pie, bar, and so on, and can compare the values of a particular data at different times. The strict test shows that the data warehouse provided by the system is reasonably designed, and can make the business and management personnel and convenient access to the desired data, can greatly improve the efficiency of the company's management, and has a better user experience. For the enterprise to provide better decision support, has good use value.
【學位授予單位】：北京工業(yè)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(guān)期刊論文 前1條

1 劉志超;陳勇;姚志立;;大數(shù)據(jù)時代的電子商務服務模式革新[J];科技管理研究;2014年01期

，

本文編號：2370938