Number of model building process cartridge

Data warehouse development has gone through three such processes:

1. 简单报表阶段：这个阶段，系统的主要目标是解决一些日常的工作中业务人员需要的报表，以及生成一些简单的能够帮助领导进行决策所
   需要的汇总数据。大部分表现形式为数据库和前端报表工具。
3. 数据集市阶段：这个阶段，主要是根据某个业务部门的需要，进行一定的数据的采集，整理，按照业务人员的需要，进行多维报表的展现，
   能够提供对特定业务指导的数据，并且能够提供特定的领导决策数据。
4. 数据仓库阶段：这个阶段，主要是按照一定的数据模型，对整个企业的数据进行采集，整理，并且能够按照各个业务部门的需要，提供
   跨部门的，完全一致的业务报表数据，能够通过数据仓库生成对对业务具有指导性的数据，同时，为领导决策提供全面的数据支持。

By stage of development of data warehouse construction, we can see that an important distinction construction building data warehouses and data marts is that support data model. Therefore, the construction of the data model for the construction of our data warehouse, has a decisive significance.

In general, the main building of the data model can help us solve some of the following questions:

• Conduct a comprehensive traffic grooming, improve business processes.
  1. In the construction phase of its business model that can help our business or management organs of the business units to conduct a comprehensive sort.
  2. Through the construction business model, we should be able to fully understand the operation of the business architecture diagram of the unit and the entire business, the business can be categorized and procedures in accordance with specific rules.
  3. At the same time, help us to further improve business processes, improve operational efficiency, to guide the production of our business units.
• The establishment of comprehensive data perspective, eliminating information silos and data discrepancies.
  1. model building data warehouse, data can provide a perspective for the whole enterprise, each department is no longer just concerned about their data.
  2. And by the construction of the model, and sketched out the intrinsic link between departments, to help eliminate the problem of information silos between the various departments.
  3. More importantly, through the construction of the data model, to ensure data consistency across the enterprise, the difference data between the various departments will be effectively addressed.
• Flexibility to address changes and data warehousing services.
  Through the construction of the data model, we can separate the good show underlying technology implementation and upper-layer services.
  When top business changes, the data model, the underlying technology can very easily complete operational changes to achieve the flexibility of the entire data warehouse system.
• Help build a data warehouse system itself.
  1. By building a data warehouse model, developers and business people can easily reach a defined range of system construction, planning and long-term goals, thereby enabling the entire project team clear the current task,
  2. speed up the construction of the whole system .

Modeling roughly divided into four stages:

• Business modeling , this part of the modeling work, mainly contains the following sections:
  1. dividing the entire business units, in accordance with the general division of the business sector, to define operations between the various parts, to sort out among business units relationship.
  2. In-depth understanding of specific business processes within the various business units and procedures.
  3. propose changes and improvements to business processes and work procedures.
  4. The scope of data modeling to define, target and phasing of the entire data warehouse project.

• Conceptual modeling field , this part of the modeling work, mainly contains the following sections:
  1. Extract the key business concepts and the abstract.
  2. The concept of packet traffic, according to business lines grouping similar concepts polymerization.
  3. The concept refinement packet, a packet clarify the concept of business processes and abstract.
  4 clarify the concept of association between a packet, a complete conceptual model field.

  Conceptual model requirements are as follows:
  

• Logical modeling , this part of the modeling, mainly includes the following sections:
  1. Business entity of concept, and considering their specific properties.
  2. The event materialized, the so-called facts, and consider the contents of its properties.
  3. Description of the entity, which is called the dimension, and consider the contents of its properties.

  Logic model requirements are as follows:
  

• Physical modeling , modeling this part, mainly contains the following sections:
  1. platform-specific physical, technical adjustments accordingly.
  2. consideration for performance model, making appropriate adjustments for a specific platform.
  3. The need for management, combined with the specific platform, and make appropriate adjustments.
  4. Generate the final execution of the script, and improve it.

  Physical model requirements are as follows:
  
  Actually Simply put, most of the above design process model is applied to DWD layer, which is the fact dimension layer. By modeling, stroke clear logic, the implementation of the business to a picture of the table, and the table to sort out the relationship between the tables.