Hi - I see some good examples and definitions from the folks. Here's my 2 cents in addition to that, the definition in general in the BI/DW world is:
- Normalization is the process of organizing your data and breaking it into smaller tables that are easier to manage. The primary reason we normalize a database is to prevent redundant data.
thanks for ur reply.I read it in an article that data normalization functionality includes conversion of lower case letters into uppercase letters.How does it come under data consolidation scenario
In <i>"Mapping Fields/Values"</i> Tab of Import Manager we may use "<b>Set Value Conversion Filters</b>" by Right Clicking on a field and use the various options from the Context Menu for Data Normalization .
Also suppose that you have data coming from various Client systems then each Client System may us different way of Storing value for <b>United States </b>like
--> Client A stores it as <b>U.S.</b>,
-->Client B Stores it as <b>U.S.A </b>and
-->Client C stores it as <b>US. </b>
Then while Importing You perform Value Mapping and Map all these to a Single Value i.e. United States in MDM repository. This is nothing but Data Normalization.
a different usage to the one Sagar has already pointed out is normalization during the Matching process. Imaginge you want to compare three customers:
1.) IBM
2.) I.B.M.
3.) International Business Machines
You can see at once that all three customers are identicals, but MDM cannot without normalizing this data! If you use the MAtching "equals" or "token equals" function on this data, you won't get any result as the strings are completely different. But you can use the normalization capability of MDM to remove the dots and to replace "International Business Machines" with its acronym. Then you'll have three times "IBM" in your matching and you'll receive a full match.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.