"A Data Warehouse System for Supporting University Research Management"
A Data Warehouse System for Supporting University Research Management
Sprache des Titels:
Decision making in university research management requires information regarding performance of the university organization units in various research related areas with respect to a certain scientific area and time. However, this information cannot be made available easily, because the data related with this information often reside in various databases and exist in a variety of formats, representations, and structures. These data have to be extracted, transformed, consolidated, and made available to the research managers for information queries to support the decision making.
This thesis is intended to examine that data warehousing can provide solution to the problem. The examination is conducted by performing a case study at the Johannes Kepler University (JKU) Research Management. In the case study, requirements related with the decision making are defined together with the JKU Vice Rector of Research. Based on these requirements a data warehouse system is developed. ETL (Extract, Transform, and Load) mechanisms are designed and implemented to retrieve and consolidate data from various sources into a DDS (Dimensional Data Store). Then, a multidimensional database (cube) is created to make the data available for queries. Finally, various queries are performed to the cube to see if the data warehouse fulfils the predefined requirements.
The case study shows that data originated from various sources can be extracted and various differences in data can be consolidated and standardized. It also shows that various operations related with information queries in support of decision making can be performed. The operations include retrieving information about the contribution of organization units to the whole performance, comparing the performance of organization units, drilling down the performance information along the organization structure, and drilling down the performance information along the research related area with respect to time and scientific areas.