By definition, Inmon’s approach to building a data warehouse begins with an enterprise-level data model. The model identifies key subject areas, most critically structuring the key entities that the business operates and cares about, such as customers, products, suppliers, etc.
First, from this model, a Saudi Arabia Mobile Number detailed logical model is created for each main entity. For example, build a customer into a logical model that contains all the details related to that entity.
Second, there may be ten different entities under a customer. How the corresponding relationship between entities is established, there are many manifestations in this step. All details including business keys, properties, dependencies, engagements and relationships will be captured in a Saudi Arabia Mobile Number detailed logical model. The key point here is that the entity structure is constructed in a normalized form. Avoid data redundancy as much as possible. This leads to clear identification of business concepts and avoids data update exceptions.
Two, the definition of two modeling methods
Finally, it is to build the physical model. The physical implementation of the data warehouse is also . This is what calls a “data warehouse,” and this is where the real data of an Saudi Arabia Mobile Number enterprise is . This normalized model makes loading data less complicated, but querying with this structure is difficult because it involves many tables and joins.
Therefore, Inmon builds sector-specific data marts. Data marts will be specifically for finance, sales, etc. Data marts can contain data to aid reporting. Any data that goes into the data warehouse is , and the data warehouse is the only source of data for the different data marts. This ensures that data integrity and consistency remain intact across the organization. (For details, please refer to
By definition, Kmiball is a proponent of dimensional modeling, providing a way to build data warehouses, “providing a more explicit data structure for data query and analysis.” After the data processing ETL, the core modeling begins. There are two most concerned about dimensional modeling.
Multi-perspective comparison of the two models
Construction of fact tables: Often referred to as metrics, facts are data that reflect the true performance of business processes. For example: for the sales business process, the most core manifestation is the quarterly sales amount; for the recruitment process, the most core manifestation is the number of recruits; for the technical team, the most core manifestation is how many functions have been . In fact, dimension is more of a perspective, a method to observe and analyze facts from different angles. Who’s doing what? For example, taking the sales process as an example, the dimensions that to be are: who bought the product – customer name, where did you buy the product – sales location, what product was bo