Three main data models are:
Approach: Records are arranged in inverted tree like structure with relationship in one direction and there can be only one parent
Disadvantages: 1. Rigid parent-child structure 2. Difficult navigation pro grammatically
E.g. IBM IMS
Approach: Records are arranged in networked node fashion. i.e. same as hierarchical with multiple parents
Disadvantages: More flexible navigation (because of bi-directional relationship)but still changes to database and application are still complicated
In both above models, structure is hard-wired in the application program, so change in database means need to change using applications as well
E.g. CODYSL DBTG
Period: Late 1970 and 1980s
Approach: parent-child relationship of above models replaced with related tables comprising of rows and columns.
Advantage: Flexibility even with complex relationships. Relational database is independent of application
E.g. DB2, Oracle, SQL Server, MySQL
It is to be noted that databases based on hierarchical and network models still exist and are in use.