The existing Penn State algorithm attempts to perform an exact match first using one of the following sets of data:
- First and Last Name (match codes), Social Security Name, and Date of Birth
- First and Last Name (match codes), PSU ID #, and Date of Birth
- First and Last Name (match codes), Userid and Date of Birth
If there were no exact matches, then based on the country specified in the address, a fuzzy match is performed using DataFlux and one of the following rankings:
If there are multiple matches, the highest ranking match is selected.
This algorithm is a scoring algorithm as opposed to a chart-based one. From what I was told by Undergraduate Admissions, they are obtaining really good results with it.