OLAP的12條準則

原文鏈接:https://blog.csdn.net/lzhat/article/details/59102150

聯機分析處理 (OLAP) 的概念最早是由關係數據庫之父E.F.Codd於1993年提出的,他同時提出了關於OLAP的12條準則。OLAP的提出引起了很大的反響,OLAP作爲一類產品同聯機事務處理 (OLTP) 明顯區分開來。
當今的數據處理大致可以分成兩大類:聯機事務處理OLTP(on-line transaction processing)、聯機分析處理OLAP(On-Line Analytical Processing)。
業務類系統主要供基層人員使用,進行一線業務操作,通常被稱爲OLTP(On-Line Transaction Processing,聯機事務處理)。
數據分析的目標則是探索並挖掘數據價值,作爲企業高層進行決策的參考,通常被稱爲OLAP(On-Line Analytical Processing,聯機分析處理)。
OLTP是傳統的關係型數據庫的主要應用,主要是基本的、日常的事務處理,例如銀行交易。
OLAP是數據倉庫系統的主要應用,支持複雜的分析操作,側重決策支持,並且提供直觀易懂的查詢結果。

————————————————
原文鏈接:https://blog.csdn.net/dancewolf2010/article/details/90746466

 

OLAP的12條準則  由關係數據庫之父E.F.Codd於1993年提出的
1: Multidimensional conceptual view OLAP  模型必須提供多維概念視圖
       User-analysts would view an enterprise as being multidimensional in nature – for example, profits could be viewed by region, product, time period, or scenario (such as actual, budget, or forecast).  Multi-dimensional data models enable more straightforward and intuitive manipulation of data by users, including "slicing and dicing".
    分析用戶能自然的視企業爲一個多維模型,例如,利潤可以按區域,產品,時間,或方案(如實際,預算或預測)查看。多維數據模型能讓用戶更直接和方便的操作數據,包括“切片和切塊”
2: Transparency   透明性
       When OLAP forms part of the users’ customary spreadsheet or graphics package, this should be transparent to the user. OLAP should be part of an open systems architecture which can be embedded in any place desired by the user without adversely affecting the functionality of the host tool. 
    The user should not be exposed to the source of the data supplied to the OLAP tool, which may be homogeneous or heterogeneous.
    當OLAP以用戶習慣的方式提供電子表格或圖形顯示時,這對用戶應該是透明的。OLAP應該是開發系統架構的一部分,這個架構能按用戶的需要嵌入到任何地方,而不會對主機工具的功能產生副作用。用戶不應該接觸到提供給OLAP工具的數據源,這些數據可能是同構的或是異構的
3: Accessibility 存取能力準則
        The OLAP tool should be capable of applying its own logical structure to access heterogeneous sources of data and perform any conversions necessary to present a coherent view to the user.  The tool (and not the user) should be concerned with where the physical data comes from.
    OLAP工具應該有能力利用自有的邏輯結構訪問異構數據源,並且進行必要的轉換以提供給用戶一個連貫的展示。是OLAP工具而不是用戶需要關心物理數據的來源
4: Consistent reporting performance 穩定的報表能力
    Performance of the OLAP tool should not suffer significantly as the number of dimensions is increased.
    OLAP工具的性能不應該因維度增加而受到明顯的影響
5: Client/server architecture 客戶/服務器體系結構
    The server component of OLAP tools should be sufficiently intelligent that the various clients can be attached with minimum effort. The server should be capable of mapping and consolidating data between disparate databases.
    OLAP工具的服務器端應該足夠的智能讓多客戶的以最小的代價連接。服務器應該有能力映射和鞏固不同數據庫的數據
6: Generic dimensionality 維度的等同性準則
    Every data dimension should be equivalent in its structure and operational capabilities.
    每個數據維度應該具有等同的結構和操作能力
7: Dynamic sparse matrix handling 動態的稀疏矩陣處理
    The OLAP server’s physical structure should have optimal sparse matrix handling.
    OLAP服務器的物理結構應能處理最優稀疏矩陣
8: Multi-user support  多用戶支持能力
    OLAP tools must provide concurrent retrieval and update access, integrity and security.
    OLAP應提供併發獲取和更新訪問,保證完整和安全的能力
9: Unrestricted cross-dimensional operations 非受限的跨維操作
    Computational facilities must allow calculation and data manipulation across any number of data dimensions, and must not restrict any relationship between data cells.
    計算設備必需允許跨數據維度的計算和數據操作,不能限制任何數據單元間的關係
10:Intuitive data manipulation 直觀的數據操縱
    Data manipulation inherent in the consolidation path, such as drilling down or zooming out, should be accomplished via direct action on the analytical model’s cells, and not require use of a menu or multiple trips across the user interface.
    數據操作應在固定的路徑下,例如鑽或縮小,應該通過直接在分析模型的單元上完成,而不需要目錄貨多次的用戶交互
11:Flexible reporting 靈活的報表生成
    Reporting facilities should present information in any way the user wants to view it.
    報表設備應該能以用戶需要的任何方式展現信息
12:Unlimited dimensions and aggregation levels. 不受限的維與聚集層次 
    The number of data dimensions supported should, to all intents and purposes, be unlimited. Each generic dimensions should enable an essentially unlimited number of user-defined aggregation levels within any given consolidation path.
    數據維度數量應該是無限的,用戶在每個通用維度上定義的聚集聚合層次應該是無限的。

發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章