外文翻译----powerdesigner的起步与数据建模(编辑修改稿)

外文翻译----powerdesigner的起步与数据建模(编辑修改稿)内容摘要：

an object representation of the CDM using the UML standard. Logical Data Models A Logical Data Model helps you design a database structure and perform some database denormalization actions independent of any specific DBMS physical requirements. You can use a logical model as an intermediary step in the database design process between the conceptual and physical designs: ♦ Start with a CDM containing entities, attributes, relationships, domains, data items and business rules. If need be, you may develop the CDM in several design steps starting from a high level model to a low level CDM ♦ Generate an LDM. Create indexes and specify FK column names and other mon features ♦ Generate one or more PDMs, each targeted to a specific DBMS implementation This design process allows you to keep everything consistent in a large development effort. Data Models A Physical Data Model (PDM) is a database design tool suitable for modeling the implementation of physical structures and data queries in a database. Depending on the type of database you want to design, you will use different types of diagrams in the PDM: ♦ Operational PDM You use PDM to design the structure of an operational database. Usually, in data modeling, the physical analysis follows the conceptual and/or logical analysis, and addresses the details of the actual physical implementation of data in a database, to suit your performance and physical constraints. ♦ Business intelligence PDM You can use a PDM to design the structure of a data environment, which consists of: • Data warehouse or data mart database – are populated with data transferred from operational databases, and gather together all the information that may be needed in an OLAP database, where queries for business analysis and decision making are performed. The data warehouse database gathers all the data manipulated in a pany for example, whereas the data mart focuses on smaller entities in the pany. You use physical diagrams to design a data warehouse or data mart database. Since these databases usually contain very large amounts of data for storage, you do not need to design them fo。