《外文翻译数据库和数据仓库.doc》由会员分享，可在线阅读，更多相关《外文翻译数据库和数据仓库.doc（11页珍藏版）》请在三一办公上搜索。
1、Database and Data WarehousesWhat is a Database?What is a database? A database is a collection of related data items. It is generally stored on secondary storage devices that allow rapid direct access to individual data items. Redundancy is minimized; where possible, only a single copy if a data item
2、 exists. The database may be used by many different application systems at once, eliminating the need for separate systems to maintain the data for each application .When a user program inquires if a particular item is in the database, a database management system (DBMS) does the actual searching. T
3、he user does not need to be familiar with the format in which the data is stored or the actual physical location of the data.The DBMS cerates the database, keeps it up-to-date, and provides reday access to authorized user. Database management systems also provide extensive security measures to preve
4、nt unauthorized access. They make it convenient for expressing relationship between related data items and facilitate the design of user application systems. They provide backup and recovery capabilities to prevent against loss or destruction of vital information. They ensure database integrity, tha
5、t is, what is supposed to be in the database is there and what is not supposed to be .isnt. a person called the database administrator determines who may access the database, modify it, add new relationships ,and the like. This is a very important lifeline. The real question in database management s
6、ystems is how to organize information to provide rapid answers to the kinds of questions users are likely to ask. The same data organized differently can yield drama-tidally different access speeds. There common ways are used to structure a database to indicate the relationships among the data items
7、; these are the relational database, the hierarchical data-base, and the network database.Hierarchical, Network, and Relational Databases (three principal logical database models)The Hierarchical Model In a hierarchical database, data records are arranged in a strict parent child relationship. Each
8、parent record may have many children, but each child record has exactly one parent. Figure 3-1 shows a simple hierarchical database, indicating the relationship between a customer and the orders it has placed with a company. Searching a hierarchical database is rapid and convenient as long as it is
9、searched from the top down. IBMs IMS (Information Management System) is the most widely used hierarchical DBMS. Hierarchical DBMS are best suited for problems that require a limited number of structure answers that can be specified in advance. Once data relationships have been specified, they cannot
10、 easily be changed without a major programming effort. Thus, the hierarchical model cannot respond flexibly to changing requests for information.The Network Model The network database model is best at representing many-to-many relationship among data. In other words, a ”child” can have more than one
11、 “parent”, as Figure3-2 illustrates. Computer Associates IDMS is a network DBMS for computer mainframes.Network DBMS are more flexible than hierarchical DBMS, but access paths must still be specified in advance. There are practical limitations to the number of links, or relationships ,that can be es
12、tablished among records. If they are too numerous, the software will not work efficiently. Neither network no hierarchical database management models can easily create new relationships among data elements or new patterns of access without major programming efforts.The Relational Model In the early
13、1970s the relational database approach was developed to provide a much more user-friendly organization. Instead of using complex structures of pointers the relational database stores information simply as tables called relations. These tables are sometimes called flat because the rows of the table r
14、eally are very much the same as the records of a file.Each row in a relation is called a record. Each column corresponds to a particular field within the record (fields are also called domains).The fields are headed by attributes, which describe the entries in the column. A relational database consi
15、sts of one or more relations. A relational DBMS performs three primary operations on relations to form new relations:(1) Two relations may be joined (combined).(2) A relational may be projected (some of the columns are extracted form the relations and used to form the columns of the new relation).(3
16、) Records may be selected according to various user specified criterias(Structured Query Language) is the most important query language based on the relational model. For instance, a relation named Accounts, record bank accounts, their balance, and type might look like: AccountsAccountBalanceType173
17、9211700.00Checking251101888.00Savings Heading the columns are the three attributes: Account No, Balance, and Type. Below the attributes are the rows, or records. The first row says that account number 173921 has a balance of one thousand and seven hundreds dollars, and it is checking account .Suppos
18、e we wanted to know the balance of account 173921.We could ask this Query in SQL as follows:SELECT balance FORM Accounts WHERE accountNo=173921IBMs DB2 and Oracle from the Oracle Corporation are examples of mainframe relational database management systems. Microsoft Access is a PC relational databas
19、e management system. (Oracle also has a PC version).Client-Server ArchitectureMany varieties of modern software use a client-server architecture, in which requests by one process (the client) are sent to another process (the server) for execution. Database systems are no exception. In the simplest c
20、lient/server architecture, the entire DBMS is a server, except for the query interfaces than interact with user and send queries or other commands across to the server. For example, relational systems generally use the SQL language for representing requests from the client to the server. The databas
21、e server then sends the answer, in the form of a table or relation, back to the client. The relationship between client and server can get more complex especially when answers are extremely large. There is also a trend to put more work in the client, since the server will be a Bottleneck if there ar
22、e many simultaneous database users.New Forms of Database and Data WarehousesObject-Oriented Database Object-oriented databases store data as objects that can be automatically retrieved and shared. Included in the object are the processing instructions to complete each database transaction. These obj
23、ects can contain various types of data, including sound, graphics, and video as well as traditional data and processing procedures. The objects can be shared and reused. These features of Object-oriented databases promises to facilitate software development through reuse and the ability to build new
24、 multimedia application that combine multiple types of data. Benefit of Odoms is their ability to support application for the World Wide Web, as described in the focus on Technology.Object-oriented database are still a relatively new technology and can be much slower than relational systems for hand
25、ing large quantities of data where there is a high volume of transaction processing. Hybrid object-relational data bases have been developed that combine the capability of handing large numbers of transactions found in relational Dams with the capability of handing complex relationships and new type
26、s of data found in Odoms.Hypermedia Database Hypermedia database manage data differently from object oriented DBMS. but they can also contain diverse types of data. They store data as “chunks”of information, with each chunk in a separate node. Each node can contain traditional numeric or character d
27、ata or wholedocuments ,software programs, graphics, and even full motion video. Each node is totally independent-the nodes are not related by a predetermined organization scheme as they are in traditional database. Insured, user establishes their own links between nodes. The relationship among nodes
28、 is less structured than in a traditional DBMS. Searching for information does not have to follow a predetermined organization scheme. Users can branch directly form one node to another in any relationship they establish. For instance, a hypermedia database on automobiles might link basic product in
29、formation with descriptive sales brochures, a video showing the automobile in action, and the location of authorized dealers.Data Warehouses Many companies have allowed their data to be stored in many separate systems that are unable to provide a console dated view of information usable company-wide
30、. One way to address this problem is to build a data warehouse.A data warehouse is a database that consolidates data extracted from various production and operational systems into one large organizations core transaction processing systems are reorganized and combined with other information, includi
31、ng historical data so that they can be used for management decision making and analysis. In most cases, the data in the data warehouse can be used for reporting-they can not be updated-so that the performance of the companys underlying operational system is not affected. The focus on problem solving
32、 describes some of the benefits companies have obtained by using data warehouses.Data dictionaryThe data dictionary (Data dictionary) is the record database which and the application procedure Yuan data table of contents one kind of user may visit. The initiative data dictionary is refers to when ma
33、kes the revision to the database or the application program structure, its content may the data dictionary which renews automatically by DBMS. The passive data dictionary is refers to when the revision must renew its content manually the data dictionary. The data dictionary is a reservation space, a
34、 database, this is uses for the store information database itself. The data dictionary possibly contains information, for example: The database design material, the storage SQL procedure, the user jurisdiction, the user statistics, in the database processs information database growth statistics, the
35、 database performance statistics, the data character canonical is in the system each kind of data description set, carries on the main achievement which the detailed data collection and the data analysis obtain. The data dictionary usually includes the data item, the construction of data, the data s
36、tream, the data storage and the treating processes five parts. The data dictionary is about the data information set, is also all element definition set which contains to the data flow graph. Composes 1 data stream, 2 data stream components, namely the data element 3 data storages 4 process the data
37、base the important part is the data dictionaries. It deposits has the related information which the database uses, to the user is group of read-only tables. The data dictionary content includes: (1) database all pattern object information, like table, view, bunch, and index and so on. Assigns how ma
38、ny spaces, current has used how many spaces and so on. (2) row default value. (3) restraint information integrity. (4) users name. the (5) user and the role the jurisdiction which awards. (6) user visit or use audit information. (7) other produce database information. the database data dictionary is
39、 group of tables and the view structure. They deposit in the SYSTEM table space. The database data dictionary is not only each database center. Moreover to each user is also the very important information. The user may use the SQL sentence to visit the database data dictionary.Data Independence, Int
40、egrity, and SecurityData Independence In database system, each program works with its own view or views of the database. If new fields are added to a database record, the DBMS preserves the existing views so that existing programs dont have to be changed. The ability to modify the structure of the d
41、atabase without affecting existing programs that refer to the database is called data independenceData Integrity Data integrity refers to the accuracy, correctness, or validity of the data in the database. In a database system, data integrity means safeguarding the data against invalid alteration or
42、 destruction. In large on-line database system, data integrity becomes a more server problem.Data Security Data security refers to the protection of a database against unauthorized or illegal access or modification. This usually involves one or more level of password protection that is specified in
43、the data dictionary. For example, a high-level password might allow a user to read from, write to, and modify the database structure, whereas a low-level password might only allow a user to read from the database.Usually an audit trail, the recorded history of the modifications to a database, can be
44、 used to identify where and when a database was tampered with and it can also be used to restore the file to its original condition. 数据库和数据仓库1.什么是数据库？数据库就是相关的数据项的集合。它通常被存储在辅助存储器上，这些设备允许快速直接地访问单个数据项。这样可以做到冗余最小，如果数据项存在的话那么它只有一个副本。数据库可被很多不同的应用系统使用，因而避免了不同的系统为每一个应用而维护一个数据库。当用户程序需要数据库中的特殊项时，数据库土管理系统（DBMS
45、）做实际的搜索。用户不需要了解数据的存储格式或数据的实际物理位置。数据库管理系统建立数据库并不断更新，为授权用户提供方便的访问方法。数据库管理系统也提供其它的安全措施以防止非授权访问。DBMS使用相关数据项之间关系的表示非常方便，使用户应用系统的设计变得容易。它们提供备份和恢复功能以防止重要的信息丢失或被破坏。由系统管理员决定谁可以访问数据库、修改数据库以及增加新的关系等。这也是非常重要的责任。数据库管理员具有对商业信息生命线最大的控制权。数据库管理系统真正的问题是如何组织信息为用户提出的各种可能的问题提供快速的答案。以不同方式组织的相同数据会产生非常不同的访问速度。为了表明数据项之间的关系，
46、通常有三种方法用于建立数据库：关系数据库、层次数据库和网状数据库。2.层次、网状和关系数据库（3个主要的逻辑数据库模型）(1) 层次模型 在层次数据库中，数据记录按严格的双亲孩子关系控制。每一个双亲记录可以有许多孩子，但是每一个孩子只能有一个双亲。图3-1所示为一个简单的层次数据库，它显示了顾客和一家公司的定货关系。从上到下搜索一个层次数据库是快捷和方便的。IBM的信息管理系统（IMS）是应用最广泛的层次数据库管理系统。层次数据库管理系统最适合于需要有限个能够提前详细说明的结构性答案的问题。一旦数据关系被详细说明，如果没有强有力的编程支持，它们是不能轻易被修改的。因此，层次模型不能灵活地满足信
47、息要求改变的情况。(2) 网状模型 网状数据库模型最适宜于表示数据之间多对多的关系。换句话说，“孩子”可以有不止一个“双亲”，如图F-2所示。Computer Associate的IDMS就是一个大型计算机的网状数据库管理系统。网状数据库管理系统比层次数据库管理系统更加的灵活，但是访问路径仍需提前说明。实际应用中，记录之间建立的连接或关系有一些限制。如果关系太多的话，软件将不能有效地工作。在不修改主程序的情况下，网状和层次数据库管理模型都不能容易的在数据元素之间建立新的关系或新的访问方式。(3) 关系模型在20世纪70年代早期发展关系数据库是为了提供更多的对用户友好的组织。关系数据库简单地将数
48、据存储为表格（这些表格叫关系）而没有采用复杂的指针结构。这些有时被称为平面文件，因为表的行同文件的记录非常相似。关系中的每一行叫一个记录。每一列是记录的一个特殊的域。域对应以域名开头，它描述了整个一列。关系数据库包含一个或多个关系。一个关系数据库管理系统对关系进行以下三种主要的操作以建立新的关系。(1) 连接两个关系（合并）。(2) 扩展一个关系（从关系中抽出某几列用于形成新的关系的列）。(3) 根据用户描述的不同标准选择记录。SQL（结构查询语言）是基于关系模型的最重要的查询语言。例如，一个叫“账目”的关系记录银行的账目，它们的结余和类型如：账号结余类型173921￥1700.00检查251
49、101￥888.00储蓄 列的头是3个域名：账号、结余、类型。域名以下是行或记录。第一行说明账号173921有1700元的结余，它是检查账号。假设我们想知道账号173921的结余，我们可以用SQL发出请求如下：SELECT balance FORM Accounts WHERE accountNo=173921IBM的DB2和Oracle公司的Oracle就是大型计算机关系数据库管理系统。Microsoft的Access是个人计算机关系数据库管理系统。（Oracle也有个人计算机版本）。3.客户/服务器结构 现代软件中有很多采用客户/服务器结构。在这种结构中一个过程（客户）的要求发送给另一个过程（服务器）去处理，数据库系统也不例外。
