Warehousing and Database Mining

Important Information That Trident University Might Obtain By Database Warehousing and Database Mining Student Records and Learning Management System Data

A Case Study of the Federal University of Technology, Akure, Nigeria

In order for an organization to remain adequately competitive in the contemporary business environment, it needs a strong foundation of high quality data. Institutions of higher learning, such as universities, need such capability just like other large companies. To enhance data quality for strategic and tactical decision making, many universities as well as colleges are developing data warehouses (Michael, 2009). Data warehouses are separate data stores extracted from production databases in order to provide authoritative support for decision making. Data mining, on the other hand, refers to “computer-assisted tools and techniques for sifting through and analyzing these vast data stores in order to find trends, patterns, and correlations that can guide decision making and increase understanding” (Phil & HodesSue, 2011). This paper reviews the successful building of data warehousing and data mining from course management systems by Nigeria’s Federal University of Technology highlighting important information that Trident University can obtain by adoption of a similar system.

The Federal University of Technology in Akure, Nigeria registers students each semester at school levels. These students undertake courses as well as exams. These processes are managed by some online transaction processing (OLTP) system which can handle only limited historical data (Akintola et al, 2011). Data warehouse development in the university was, however, necessitated by its need for a strategic decision making support. Furthermore, other systems such as the Unified Modeling Language (UML) and the Entity Relationship (ER) handle data that are not sufficient for making strategic decisions (John & John, 2010). This data warehouse, together with the OLAP, has enabled both management and staff to perform roll-up as well as drill-down operations on student enrollments based on year, semester or any other desired combination. The new system also provides a strong, user-friendly and flexible decision support. These provide a basis upon which Trident University, with more or less similar operations, can reflect the need for data warehousing.

The Federal University of Technology, Nigeria also makes use of data warehousing and data mining in answering questions like: What are the emerging admission and students’ enrollment trends in the university. The system also facilitates schedule and allocation of Lecture rooms based on annual enrollment (Akintola et al, 2011). In order to answer these questions, there is great need for a large pool of historical support data which could not be generated by the former OLTP system. First, it is important to establish what kinds of data will need to be assembled in order to develop a working data warehouse and data mining system (Phil & HodesSue, 2011). Other Universities that have also implemented the same system such as Arizona State University have had to address numerous data issues in the building process of their data warehouses. These include determination of what kind of data to gather, the frequency of updating the data, ensuring that the data remains official and resolution of issues of data privacy and data security. ASU’s experience indicates that it may not be possible to completely address these issues with a data warehouse (Michael, 2009).

Data warehouses must provide the very right data and to the very right people (Akintola et al, 2011). At the FUTA, after selecting the star schema for use, dimensional tables were collected containing data such as student identification, faculty identification, classroom identification, class identification as well as course identification. Each of these dimensional tables contained various attributes and measures such as individuals’ bio-data (John & John, 2010). However, it is important to note that it may not be possible to deliver all the data which people want. People are bound to asking different questions over time hence it is not easy to predict what kind of data they exactly need (Michael, 2009). ASU began by asking users which data they exactly wanted and the kinds of reports they utilized. Another good strategy is start by looking at data that goes to the research department of the institution. ASU’s experience shows that once implemented, users of the data warehouse quickly make known to the development team the kind of data that they want (Akintola et al, 2011). Trident University can therefore begin from the basic data that is available at the school level moving across the departments with time.

From the experience of FUTA and other institutions such as ASU, it is evident that data mining the data stores in the data warehouse can facilitate strategic decision making. Questions of lecture room scheduling and allocation can easily be achieved despite the large population of the university. This is because the data warehouse contains dimensional tables with attributes on all lecture room identification and timing of respective lectures. This will help a great deal in minimizing duplication of lecture schedules in same classrooms as was the case before the system was adopted (John & John, 2010). The management also finds ease in coordinating the departments since they can easily extract the number of students enrolled in a particular major. Planning and resource allocation for university examinations across all the divisions can be easily done. It is also possible to tell which students qualify for examinations based on set thresholds such as lecture attendance, fees payment and registration (Akintola et al, 2011).

Once the required data is collected, implementation of the system will begin by transporting all the data from the OLTP database to the warehouse. This can be done by use of the SSIS tool provided by the SQL Server. Data structure is created for the dimension tables, after which an integration service project can be created using by the Business Intelligence packages of the SQL server. Project creation is followed by creation of a data-source, the OLEDB data source as well as the destination. The dimensional tables as well as fact tables are created in the process of project creation (Michael, 2009). After populating the data warehouse with data, the next stage is providing the users tools for data analysis such as the OLAP (Online Analytical Process) as well Data Mining for analysis of the process to facilitate decision making. Roll-up as well as drill-down processes of OLAP generate data analysis reports as output for various dimensional hierarchies (Akintola et al, 2011). Trident University may endeavor to understand this procedure before embarking on implementing the system.

In addition, the data warehouse can provide data analysis reports based on various aspects or question. In the University setting the following have been achieved: analysis reports on the basis of student gender as well as courses registered, Registration based on academic year data, Analysis on the basis of academic year and registered courses, students lecture halls based on time-block to show the numbers of students per courses. Reports of analysis on the basis of time tables and lecture halls against registered courses can also be obtained (John & John, 2010). With data mining, the computer can search for various correlations within the data then present significant hypotheses for users’ consideration. The problems which will be solved through data mining include: Categorization- classifying sets of cases, Clustering – natural grouping of sets of cases, Association rule – establishing items which are often processed together and Data Clustering. Discriminate mining which shows students’ marks-faculty relationships further enhance the grading system (Michael, 2009).

References

Akintola, K.G., Adetunmbi, A.O., & Adeola, O.S. (2011). Building Data Warehousing And Data Mining From Course Management Systems: A Case Study of Federal University of Technology (FUTA) Course Management Information Systems. Information Technology for People-Centred Development (ITePED 2011). Retrieved December 20, 2011, from http://www.ncs.org.ng/wp-content/uploads/2011/08/ITePED2011-Paper4.pdf

John, D.P., & John, J.R. (2010). Lessons from a Successful DataWarehouse Implementation: A case study of Arizona States University. Journal of Information Technology Education by Arizona States University. 5(26), 113-123.

Michael, A.K. (2009). A Realistic Data Warehouse Project: An Integration of Microsoft Access and Microsoft Excel Advanced Features and Skills. Journal of Information Technology Education. 8(21), 25-29.

Phil, R., & HodesSue, H. (2011). Data Warehousing at Baylor University: The Second Year. Baylor,MA: Baylor University Waco, TX.