With a process-driven approach, you define your functions and programs first and try to fit the data required into the processes. With a data-driven approach, you begin with defining the data first. Each data element is defined once and only once.
A data-driven approach eliminates data redundancy and sets the stage for normalizing the database, making it easier to be accessed and maintained.
Defining a data model helps insulate the data structure from the process logic. That is, when process logic or flow changes, the data structure or model does not need to change. Since process logic is more susceptible to change than the data model, the structure is easier to maintain. Using an Order Entry example: if the Order information is defined by the order entry transaction, the database is likely to change when the transaction changes. On the other hand, if the order information is defined based on all the information that users think of regarding an order, the data structures are likely to accommodate any set of transactions.
Your data model restricts the number of allowable operations and suggests processes that can operate on the data. Some database functions are common to all entities regardless of their structure. For example, editing a Customer entity is the same process as editing a Supplier entity although they have different entity attributes. Inquiry and reporting processes are possible for all entities.
This set of processes leads to process structures that can become building blocks for larger, more complex processes. By reducing or eliminating redundancy, processes are easier to manage and maintain.
|
Copyright © 2014 CA.
All rights reserved.
|
|