In database management systems, partial dependency is a functional dependency that refers to the phenomenon where a primary key determines the outcome of another attribute or set of attributes. It occurs when a non-key attribute of a table in a database is dependent on the value of only a part of the table’s primary key but not the entire primary key.
Functional dependency is also referred to as normalization, and it was introduced in 1970 by Professor Codd. It defines three normal forms (first, second and third) and is used to avoid three anomaly types that may affect databases. These three anomalies are insertion, deletion and update. An anomaly occurs when a database is incomplete, inconsistent and/or contradictory. The basic aim of normalization is to create a proper description of the attributes of particular data. It highlights relationships in different data in a table. For example, a table of students may have student name, course, admission number and grades as the fields of entry. If one or more of these fields are related, normalization ensures that any change to the details of one student does not affect the details of another. With anomalies, user cannot insert new data even when it is absolutely necessary to do so. In addition, deleting the data of one student may result in the loss of data of another. Similarly, when a record is updated, other instances of the same item are not updated.