Data Quality. How to analyze it?

The database can store incorrect data. It is fact. Even all check constraints passed the value can be wrong. In general, for "data quality" we mean a number of errors in the table or column. There are samples of incorrectness:

• Value has incorrect format or presentation: NNN NNNNNNN for a phone number instead of expected (NNN) NNN-NNNN.
• Value violates complex dependency: hire date have to be less than a birth date for at least 16 years.
• Value has wrong capitalization: AaAAa instead of Aaaaa.
• A number of depended items greater than required.

DTM Data Scrubber is designed to find data problems. It is two step solution: the user describes correct data state. The software analyzes existing data items and creates HTML (sample), Excel or text data quality report with information about wrong value, unique key of the wrong row and row number.

Download free demo version (?).

DTM Data Scrubber can help to find problems at the database level. For example, some "dictionary" table must contain at least 25 unique values. The metadata problems like incorrect trigger behavior can be also detected.

See Also

• How does the data cleaning process work?