Normalization is a design technique that is mostly used as a guide in designing relational databases. Normalization is necessary a two step process that gives data into tabular form by removing repeating groups and then replaces duplicated data from the relational tables.
Redundant data wastes disk space and builds maintenance problems. If data that allows in more than one place must be modified, the data must be modified in exactly the same way in all locations which is prone to errors and time consuming. A modification to a customer address is much simpler to do if that data is stored only in the Customers table and now in the database.
Inconsistent dependency is a database structure that makes certain assumptions about the location of data. As, while it is intuitive for a user to look in the Customers table for the address of a particular client, it may not create sense to look there for the salary of the employee who invokes on that customer. The employee's salary is exist to, or dependent on, the employee and thus could be placed to the Employees table. Inconsistent dependencies can create data difficult to access because the path to find the data cannot be logical, or may be broken or missing.