View Source

Structure of the University
- Departments, Divisions, Groups
- Different organizations within the campus might have conflicting views of this data
  - For example, a research group that reports to one unit but is funded by another.
Names and network IDs from an LDAP directory
Personnel records from Human Resources
- Must be filtered to remove sensitive data.
Grants information from the Office of Sponsored Funds
- Because of data challenges, this data is only harvested a few times each year.
List of classes from the Registrar's Office
Publications from a faculty reporting system
- This also provides many challenges.
  - For example, two co-authors may provide slightly different names for the same journal article, or for the journal itself.
- If the data comes from a well-curated source (e.g. PubMed), the challenges of data cleaning are greatly reduced, but the challenges of disambiguation are increased.
  - In the on-campus faculty reporting system, it is common practice to use an ID that definitively associates the data with the faculty member.
  - When reading from a public source, are these the same authors or different authors?
    - James L. Fox
    - J. L. Fox
    - J. Leroy Fox
    - James L. Foxx