Two standard methods of organising data for data processing applications are file oriented approach and database oriented approach. File oriented approach was traditional method used in early days of data processing. It has been replaced today largely by database oriented approach. However, several applications, dealing with simple and small data sets, use file oriented approach, even today.
Hence, both these methods are described below.
File oriented approach:
This method organizes an application's data into one or more files and the application program processes the data stored in these files to generate desired output. For example, it is customary to use a master file of permanent ( and, usually, the latest) data, and transaction files containing data of temporary nature. For instance, in a pay roll application, in additional to containing all permanent details of each employee ( name, address, employee code, pay rate, income tax rate, etc.) master payroll file also contains current gross-pay-to-date total and tax-paid-to-date total. Transaction payroll file contains details of hours of worked this month, normal and overtime hours, and if peace work is involved, quantity of goods made. When the user runs monthly cycle of payroll application, the application uses both master and transaction files to generate this month's payslips, and updates the master file to make it ready for following month's processing.
Database oriented approach:
File oriented approach of Organising data for data processing applications is simple, inexpensive, and usually easy to use. It, however, suffers from following limitations:
1. Limited query flexibility:
File oriented approach organises records in a file according to a key field. Hence, it can quickly process queries based on key field value. However, when the key field is not relevant to desired information, it needs to search entire file. For example, if the key field is employee code in an employee file, and if a user needs to list names of all employees having certain educational background, the application has to search all records to produce this information. If the user needs this information frequently, he/she needs to create a new file structure on educational background key with a new program to process this file.
2. Data redundancy:
In file oriented approach, same data items are often present in many different files. For instance, in the example above of creating a new file structured on educational background key, the new file will also have several data items present in the first employee file. Similarly, let us assume that a bank uses file oriented approach for storing in costumers data, and it maintains separate costumer files for its various types of services, such as saving account, term deposits, and loan accounts. In this case, many data items for a bank costumer ( such as home address, age, occupation, and identification information) may be present in multiple files. Repetition of same data items in more than one file is known as data redundancy. It leads to increase in cost of data entry and data storage. It also needs to load data integrity problem described below.
3. Data integrity problem:
Data integrity refers to consistency of data in all files. For consistency, the system must carry out any change in the data item in every file containing that data item. For instance, in our example of a bank maintaining separate costumer files for each type of account, when a costumer moves to a new address, the bank must update his/her address field in all costumer files containing this costumers regard. Integrity of data is necessary to avoid confusions that may result when an application changes a data item, if it doesn't update the data item in all files containing the data item, there may be frequent discrepancies among reports, which the application produces from different files, causing confusion.
4. Lack of program/Data independence:
In file oriented approach, an application program usually contains data format statements, which define the format of each data field precisely as the application need it for processing. Because different applications create different files, this often results in different files having same data item stored using different data formats. Data dependence occurs when data is dependent on application. Due to data dependence problem, whenever there is a need to add, delete and change data formats, we also need to change application programs. Data dependence problem also causes incompatible among data files from different applications due to which we can not link this data files, if such as need arises.
5. Limited data security flexibility:
Normally a file oriented approach offers file level data security feature. Hence, it can enforce data access restrictions on an entire file only, not on a second or a field of data item.
Dissatisfied with these limitations of file oriented approach, researchers began lot for a better method of organising data to consolidate activities. Their efforts resulted in data base oriented approach for organising data. This approach integrates together data from multiple related files in the form of a data base having following properties:
1. It provides greater query flexibility.
2. It reduces data redundancy.
3. It solves data integrity problem.
4. It makes data independent of application programs, enabling users to add, change, and delete fields from data base without effecting existing programs.
5. It also includes data security features at data base level, record level and even at field level to provide greater flexibility of restricted data access.
Comments
Post a Comment