Constructing a student data warehouse

The University of Pittsburgh currently employs a student system established more than a decade ago using SCT ISIS, which has been highly customized over the years. This mainframe legacy system lacks the flexibility to provide ad hoc reporting or web-enabling of queries and reports. Many University schools, departments, and programs require customized reports and frequently maintain their own shadow systems in order to be able to flexibly query data and produce ad hoc reports. The number of queries and reports that are maintained both centrally and in shadow systems, coupled with the frequency in which changes are needed clearly signals the need for a stable and reliable ad hoc query and reporting system.Attempts to develop a student data warehouse over the past five years have been largely unsuccessful for three reasons: the inability to reach consensus on a security model, a data model, and the difficulty in identifying an appropriate query tool. These problems were overcome through the implementation of a cooperative strategy between developers and stakeholders in which business requirements and functional specifications were clearly developed prior to the completion of a design specification document.The assessment process underlined the need for a system in which all available data elements from the current student system are available for query and that the system is sufficiently flexible to allow for general "canned" queries and reports and to permit users to employ a data query tool to design their own queries and reports. Skilled users will have the ability to employ other data mining tools, such as SAS or even Microsoft Access to query and analyze data via SQL query processes.This presentation will describe the development process in detail and examine the results of the current development efforts, still ongoing at the time of this writing.