2. Motivation For This Project


Contents:
2.1 Decision-making Processes
2.2 Integrating Multiple Data Sources and Platforms
2.3 IT Backlog
2.4 End User Data Access
2.5 User Friendly Data Design


2.1 Decision-making Processes

Management processes at Avondale College have tended not to rely to a major extent on input from computer-based information systems. There are a number of reasons for this.

  1. The financial software was developed in-house almost twenty years ago, and is primarily a record keeping system. The system does not produce management information.
  2. Administrators and decision makers have not worked with systems that provide decision support, so they are not experienced with the kind of assistance that state-of-the-art management information systems can provide. As a result, their requests for information have been infrequent and relatively unsophisticated.
  3. When the Computer Services Centre was asked for decision support information, it was usually not able to provide it in a timely manner with the current systems, and that discouraged the administrators from making further requests.

2.2 Multiple Data Sources and Platforms

One of the major difficulties faced by decision makers and knowledge workers is that the corporate data resides in many different systems. Data is held in many different formats and on a variety of platforms.

The major information system components are as follows:

Component Platform Data Format
Student Administration VAX indexed flat files
General Ledger VAX Databus emulation
Budgeting, Financial Planning PC spreadsheets
DEET Reporting
* Student
* Staff
* Finance
.
VAX
none
VAX
.
indexed flat files
manual
Databus Emulation
Banking VAX Databus emulation
Payroll PC Progress
Human Resources PC dBase
Accounts Payable none manual
Accounts Receivable VAX Databus emulation
Asset Management VAX Databus emulation
Recruiting, Marketing & Promotion PC Microsoft Access
Management Information none none
Ad hoc enquiry/analysis none none
EIS PC spreadsheets
Other none hardcopy
Table 1. Information System Components and Platforms

The same data item sometimes has different names or different attributes in each system. A simple example is the different ways that student ID can be treated in different systems.

STUDENT_ID                        PIC 9(10).
STUDENT_NUMBER                    PIC 9(5).
NAME_NUMBER                       PIC X(6).
    (NAME_NUMBER replaces STUDENT_NUMBER plus CHECK_DIGIT)
Table 2. Semantic Differences in STUDENT_ID

2.3 IT Backlog

Because the data is not in a form suitable for user-generated analysis, enquiry and reporting, each request for decision support information requires special routines to be written by programmers. The IT department has a large backlog of projects and requests typical of the industry. These include requests for modifications, bug fixes, additional functionality and enhancements to existing systems, as well as new projects for replacing legacy systems and adding new systems, either through evaluating and implementing packaged software or in-house development.

The demands for accurate and timely management information to support ad hoc enquiries and reports, graphical and what-if analysis are increasing, and the IT backlog continues to grow larger. The answers to the questions are held in the data, but it is too difficult and time consuming to extract the information needed to find the answers. This results in questions not being asked, and in duplication of personal information systems, with inadequate controls over data integrity and accuracy.

2.4 End User Data Access

One solution to the problems caused by the IT backlog is to put user-oriented reporting and analysis tools in the hands of the users. This will assist the decision making process and it will help to realize more of the potential benefit of information systems in meeting the corporate goals.

The problem is that in its current format, corporate data is not accessible via end-user tools.

  1. All of the financial data is processed using Databus programs via an emulator on the VAX. Software systems outside the emulator cannot access the financial data.
  2. Student Administration data is accessible to other applications running on the VAX, but it is not held in a relational database. Hence, the SQL server used by most client/server EIS tools cannot access student data.
  3. There is only limited personnel information in the Payroll system, but that is on a PC so even that limited information is not available to users on terminals or PC’s networked to the VAX.

The long-term objective for Information Systems (known as the Administrative Software Project) is to develop and maintain a central corporate database in a form conducive to analysis via EIS tools. However, the problem here is that the timescale for completion is 5 to 8 years, so this does not address the urgent need for improved management information.

There is a need to:

2.5 User Friendly Data Design

In order to provide end user access to data, users need to find it easy to navigate and use the database. Rather than wait until a fully integrated package can be purchased or developed in-house, a database can be designed to satisfy the requirements for ease-of-use by end users, with routines to populate it from existing data sources. This is essentially the concept of the Data Warehouse. Since this database will not be supporting operational data processing, its design can be optimized to support the needs of decision makers and knowledge workers.


Next section, Previous section or return to table of contents


Last Modified: Monday, September 2, 2002 12:12 PM