This article comes to us from Michael K. Campbell. Michael writes "Humans can instantly spot the difference between "411 Madison Avenue" and "411 Madisan Av". Moreover, because human brains are wired semantically, they can discern that both addresses are likely the same even though the second one has been misspelled. Computers, unfortunately, are typically unaware of these semantic similarities, which can lead to improperly checked data." He then goes on to show how the Fuzzy Lookup Transformation in SQL Server Integration Services (SSIS) can be used to make this exact same match. And he includes the application so you can test it yourself.
Thanks for the Article. Very helpful. Just one question. My reference table is very large. I assume indexes are used to find exact matches. I notice that the package progress says taking a snapshot of the reference table. I assume this is building some sort of hash table. Are indexes on the either table when the Fuzzy logic part is running?
Article: Using Fuzzy Lookup Transformations in SQL Server Integration Services
Posted - 01/22/2007 : 08:19:32
