Applying data warehouse in education management and decision support making process

APPLYING DATA WAREHOUSE IN EDUCATION MANAGEMENT AND DECISION SUPPORT MAKING PROCESS Ho Cam Ha and Pham Xuan Hung Hanoi National University of Education Abstract. The data of education and training system, which have been and are being gath- ered, are implicit for allowing the use of OLAP and Data Mining. In order to implement these applications, it is needed to construct a data warehouse. Based on that foundation, we will quickly and effectively have much of useful information in study, statistics and discovery. This information will be used to advise and help Managers in decision making. In our report, the implementation of a data warehouse for a pedagogic university and its effects in education are presented. 1 Introduction The application of Information Technology in Education Management in our country has been experienced for several years. It is possible to see some following features of the existing computer-based and data-based education management reality. (1) Data is collected and saved in many places and has been accumulated for many years. This data contains much useful information. However, they only mainly serve the operational information systems of each unit. (2) The applied softwares used for management of units, organizations and manage- ment levels have been developped and put into using spontaneously without a common standard. The Information Technology knowledge of officers and managers are still lim- ited. Therefore, statistics and data processing in management do not follow an optimal close procedure, a mixture of hand-made and computer-made make it difficult to detect any mistakes or errors. (3) Data is not synchronous between different units, localities and levels, i.e they are not homogeneous in terms of kinds and formats. Most of the data is not kept as archives in a scientific and consistent way. (4) Qualities and plans of training in pedagogic universities are closely and directly related to training qualities in schools. Conversely, those entering pedagogic universities are also students graduating from schools. Yet, the data of education management at schools and that at pedagogics universities are, to date, still separated. This can be seen as an unreasonable thing leading to a waste of information. In the above situation of the data management in education and training, to improve the quality of information for a better management in pedagogic universities, this study is aimed at designing a data warehouse for pedagogic universities based on some subjects. The construction of such a subject-based data warehouses is a solution for integrating the data sources, which have been and will be available, to raise the quality of information exploited from these [1]. Nowadays, the software tools for data integration, for establishing 1 and exploiting data warehouse have appreared in the common sets of software for database management such as Oracle, MS SQL Server, IBM, etc. This is a technological advantage for building and exploiting the data warehouse to serve the management tasks in the pedagogic universities effectively. 2 Data warehouse and DW technology What is a data warehouse ? Data warehouse is a subject-orriented, integrated, time- variant and nonvolatile collection of data in support of management's decision-making process [2]. Thus, it is possible to build data warehouses (DW) from the operational database in use. When using the DW Technology, comparing to the operational database, some following results can be made: (1) Data from different sources are synchronized and integrated. (2) Data is multi-dimensional. (3) OLAP services can be used. Complicated statistics which are meaningful and highly-visual can be made quickly and automatically. (4) The services for finding and exploiting data can be used. From a big volume of data stored for a long time (maybe for some years), it is possible to discover rules and to detect unusual cases, to keep track of the evolution and to assess or to forecast trends, etc [3]. 3 Findings The authors have investigated the situation of data collection and processing in some Departments of Education and Training and at the Hanoi University of Education. After analyzing and selecting subjects, the authors designed a pilot data warehouse for the subject of regular training of the Hanoi University of Education. The data were taken from the operational database of the university, which is in the form of Access tables collected from the department of training. After initial processing to make the data fresh, MS SQL Server 2005 was used to integrate sources, to form the data warehouse and the online analysis tool of SQL Server 2005 was used to print out reports and design special queries. Some main findings are presented below. 3.1 Designing diagram and procedure of building data warehouse 2 Figure 1. Diagram of measurement-dimension relation (NCKH) and diagram of measurement-dimension relation (students) Figure 2. Structure of data diagrams in data warehouse 3.2 Special queries used for analysis and decision making support With the application of the DW Technology, after experiments, it is possible to find some meaningful information in analysis and decision making consultancy 1. Assess training qualities of a course (based on faculty) by a series of reports show- ing different aspects based on different norms, for example analyzing and comparing the entrance marks between faculties, between courses in the same faculty; making charts for monitoring and comparing GPA of each discipline through semesters of many courses; graduating results (classified) between faculties and between courses compared to the re- 3 Figure 3. Procedure of building data warehouse spective entrance results ; charts for monitoring science research results, etc. 2. Assess working qualities of pedagogic students after graduating, based on that to assess training qualities of faculties in a pedagogic university. The database of Depart- ments of Education and Training has information about teaching staff (origins, qualifica- tions, training courses attended), and information about the students' studying results (by schools, districts and provinces). Once the data of the pedagogic universities are integrated to that of Department of Education, with the DW Technology, we will have information for assessing working qualities of different courses of pedagogic students after graduating. 3. Forecast needs in quantity and quality of school teachers in coming years, supporting leaders in making decisions whether or not to open more training courses for teachers, in which provinces, for which disciplines, and reasonable numbers in consideration of the future needs. These forecasts can be made from the data of number of students of districts and provinces, and the data of school teaching staff (by schools, districts, provinces) which are available at the Department of Education and which are integrated to the data of training at the pedagogic universities. Especially, it can be said that the statistic qualities using data warehouse are much better than the current statistics made on the working data. This comparision can be seen more concretely in the table below. Table 1. Comparing statistics made on working data to statistics made on data warehouse Current situation of statistics Statistics using data warehouse Semi-manual, from many different sources of data in many places, inconsistent Integrated automatically from many sources of data, automatically made in statistics based on subjects Time-consuming Less time-consuming Often in the form of tables, not visual Customized with high visuality Not multi-dimensional Multi-dimensional information Unhistorical Can be monitored throughout historical process Figure 4. Monitoring studying results of K53 with 2 codes of discipline in a faculty 4 Figure 5. Statistics by faculties of the number of students in different provinces admitted to the Hanoi National University of Education (K53), detailed by localities 4 Conclusion On the existing operational database, the construction of data warehouse is feasible and brings important benefits to the education management of pedagogic universities. With data warehouse, it is possible to exploit a huge volume of data, which is accumulated for many years to get information of high quality, which helps make strategic decisions for the universities. Therefore, it is needed to have studies to design and build subject- based data warehouses on the available database from working information systems of the universities. REFERENCES [1] Efraim Turban and Jay E. Aronson, 1998. Decision Support Systems and Intelligent Systems , Prentice - Hall International Inc. [2] Jiawei Han and Micheline Kamber, 2000. Data Mining: Concepts and Techniques, Prentice - Hall International Inc. [3] A. Berson, S J. Smith, 1997. Data Warehousing, Data Mining, & OLAP, McGraw Hill. [4] Ho Thuan, Ho Cam Ha, 2004. Database Systems: Theory and practice. Vol. 1, Vol.2, Educational Publishing House. [5] J. Bischoff & T. Alexander, 2002. Data Warehouse: Practical Advice from the Experts, Prentice Hall. [6] V. Poe, 1996. Building a data warehouse for Decision Support, Prentice Hall. 5