News Update!
  In Fall, 2007, a
  new Architecture &
  Standards committee
  has begun meeting

WMU > Strategic Plan

Data Warehouse Proposal - November 29, 2000

• Purpose
• Goals
• Scope
• Considerations
• Next Steps
• Conclusion
• Team Members

Currently, information is stored on multiple platforms and in many formats throughout the university without the ability to easily access and analyze the various data. However, there exists an ever-increasing need for university decision-makers to have this information at their fingertips. In addition, the Data Administration and Data Warehouse subcommittee (DADW) recognizes and promotes the importance of data as a valuable university resource. For these reasons, DADW continues to support the development of a campus-wide data warehouse.

Purpose

The data warehouse will present to university decision-makers information based on data that is accurate and consistent. As the data elements of the warehouse will be captured at pre-determined strategic times (i.e., beginning of semester, census, after grade run), the information presented through the reporting function will remain consistent rather than changing on a continual basis as it now does when extracted from a transactional system.

[Top]

Goals

While the ultimate goal of the data warehouse will be to enhance the use of data resources in support of university decision-making and strategic planning, more specific aims are listed below.

• Create a data architecture that supports the informational needs, as well as the business functions, of the university.
• Improve data management and access through the use of appropriate methods, tools, and techniques.''
  • Promote data consistency and standardization throughout the university.
  • Improve the quality, accuracy, and integrity of university data resources.
  • Minimize duplication in capturing, storing, and maintaining data.
• Encourage and facilitate data access and data sharing by providing a user-friendly method of extracting needed information.
  • Standard (canned) reports that can be generated by the end-user.
  • Query-based report-generation tool for easy development of ad-hoc reports by end-users.

[Top]

Scope

Eventually, the data warehouse will encompass data gathered from all areas on campus including Student Financial Aid, Student Admissions, Student Financial Accounting, University Libraries, Faculty Course Data, Human Resources, and the Office of the Registrar. However, the warehouse must be built in sections (or stores), as it would be impossible to encompass all of this information at once.Because of this, the DADW recommends that the warehouse begin with data contained in the student information system (academic records). The data element chart (see Appendix 1) lists those data fields that would be included in the initial development phase.

[Top]

Considerations

The primary consideration in development of the data warehouse is cost. While it is too early in the planning process to have specific numbers, development costs are likely to be incurred in areas including, but not limited to, dedicated hardware, specific software, appropriate staffing (i.e. database manager and data analyst) and training for technical support staff as well as the end-user community. Some of these costs will be one-time expenditures, while others will be ongoing.While further research is needed, the DADW anticipates that the hardware platform will be a client/server environment utilizing a Unix or Novell operating system. There are several software vendors being considered. The working list includes:

• Acta Technologies, Inc.
• Ardent Software
• Computer Associates
• D2K
• Evolutionary Technologies, Inc. (ETI)
• Informatica Corporation
• SAS Institute

[Top]

Next Steps

To get executive approval, a budget must be developed. In order to develop a meaningful budget, the hardware and software components need to be identified. From that basis, a budget can be projected with more accuracy. Once approved, hardware and software can be purchased. Upon project approval, work can begin on data analysis, and data extraction policies can be developed. This step can be completed prior to receiving the hardware and software.Final steps will be to develop data extraction instruments, populate the warehouse, develop reporting tools, test the initial data, and then release the data to the end-users. Note that the steps beginning with data analysis will be repeated for each data store as the information sources throughout the university are incorporated into the warehouse.Each of the steps indicated will involve several sub tasks. This will require cooperative efforts and involvement of many individuals on campus. For example, each data custodian will be intimately involved in the analysis of the data maintained by that office.

Conclusion

While the data warehouse project is extensive, the DADW believes it is critical to meeting the growing informational needs of the university community, especially those charged with making decisions and with strategic planning.