Data is everywhere now. Every business needs data, runs on it, and decides the current and future strategy for business by analyzing historical data. Data is very important, but not all data/information is necessary or relevant to a particular business. So, not all data should be used or stored by the system. One has to identify the data that is important to business, store it, and then provide knowledge, using it for future reference. Once we have identified the data that is important, how we should store it? Should we do it in the raw format or define a structure to store that data? To make sense of and gain knowledge data, it has to be stored in a structured format. The process of identifying and defining this structure is data modeling.

According to Wikipedia, a data model is:

Data modeling is a visual representation of organizational data semantics. It includes entities involved in businesses, such as customers, products, orders, and so on. It also includes the relations between them. Entities are the main factors to store in the database, along with their attributes. For example, for customer attributes, there are customer name, address, phone number, and so on. Another thing to consider is cardinality, which represents the degree of relationship. For instance, a customer can order multiple products and also place multiple orders. So, the cardinality between the customer and their orders is one-to-many. Many-to-one and many-to-many are other forms of cardinality.

In other words, we can rephrase the sentence as, a data model is a relatively simple representation of complex real-world data. At every level, the user requires different views of data. So, a good data model should be capable of fulfilling the needs of all different users with different permissions.