This research investigates the practical issues surrounding the development and
implementation of Decision Support Systems (DSS). The research describes the traditional
development approaches analyzing their drawbacks and introduces a new DSS development
methodology. The proposed DSS methodology is based upon four modules; needs' analysis,
data warehouse (DW), knowledge discovery in database (KDD), and a DSS module.
The proposed DSS methodology is applied to and evaluated using the admission and
registration functions in Egyptian Universities. The research investigates the organizational
requirements that are required to underpin these functions in Egyptian Universities. These
requirements have been identified following an in-depth survey of the recruitment process in
the Egyptian Universities. This survey employed a multi-part admission and registration DSS
questionnaire (ARDSSQ) to identify the required data sources together with the likely users
and their information needs. The questionnaire was sent to senior managers within the
Egyptian Universities (both private and government) with responsibility for student
recruitment, in particular admission and registration.
Further, access to a large database has allowed the evaluation of the practical suitability of
using a data warehouse structure and knowledge management tools within the decision
making framework. 1600 students' records have been analyzed to explore the KDD process,
and another 2000 records have been used to build and test the data mining techniques within
the KDD process.
Moreover, the research has analyzed the key characteristics of data warehouses and explored
the advantages and disadvantages of such data structures. This evaluation has been used to
build a data warehouse for the Egyptian Universities that handle their admission and
registration related archival data. The decision makers' potential benefits of the data
warehouse within the student recruitment process will be explored.
The design of the proposed admission and registration DSS (ARDSS) will be developed and
tested using Cool: Gen (5.0) CASE tools by Computer Associates (CA), connected to a MSSQL
Server (6.5), in a Windows NT (4.0) environment. Crystal Reports (4.6) by Seagate will
be used as a report generation tool. CLUST AN Graphics (5.0) by CLUST AN software will
also be used as a clustering package.
Finally, the contribution of this research is found in the following areas:
A new DSS development methodology;
The development and validation of a new research questionnaire (i.e. ARDSSQ);
The development of the admission and registration data warehouse;
The evaluation and use of cluster analysis proximities and techniques in the KDD process
to find knowledge in the students' records;
And the development of the ARDSS software that encompasses the advantages of the
KDD and DW and submitting these advantages to the senior admission and registration
managers in the Egyptian Universities.
The ARDSS software could be adjusted for usage in different countries for the same purpose,
it is also scalable to handle new decision situations and can be integrated with other systems.
Date of Award | 2001 |
---|
Original language | English |
---|
Awarding Institution | |
---|
BUILDING DSS USING KNOWLEDGE DISCOVERY IN DATABASE APPLIED TO ADMISSION & REGISTRATION FUNCTIONS
EL-RAGAL, A. A. H. H. (Author). 2001
Student thesis: PhD