SSIS – Fuzzy Lookups

With the any kind of database import there are always going to be data problems. Why? Well, it generally down to human beings. A computer is logical, a human is not. A computer does what its told, a human does not. So is it no wonder why data that you import which was keyed in by humans has errors? Its understandable, but what can you do about. Lets take a specific example, you have a list of employees sooner or later there are going to be spelling mistakes or typos. Other problems could be categorising, I could think the “sales” department someone else could think “Internet sales” department. So what do you do when your trying to standardise data? A Lookup transformation task wont cut it because Lookups are searching for exact matches, this is where the fuzzy lookup comes in.

The fuzzy lookup does not “look” for specifics, instead they use patern matching to best guess a solution. This tool is valuable when coming to produce your data warehouse.

This video shows you how to configure a Fuzzy Lookup and what you can do with the results returned. To fully understand Fuzzies, you must first understand Lookups (see related video). As before, this area warrants over an hour of coverage, so to save time I have produced an example database with a package already created that gives you the ability to play around to see all the different aspects of Fuzzies.

To obtain the sample files please go to my website. Happy importing!