DAT Data Quality

Question 1

Define Data Profiling

Answer 1

The process of doing the initial data assessment for a data set.

Question 2

What are the steps in Data Profiling?

Answer 2

1. Individual fields - Check content against data/domain definitions. Count violations and nulls.
2. Tables - Check inter-field relationships and keys. Check inter-table relationships.
3. Check whole data set - Are business rules fulfilled?

Question 3

What is Data Quality?

Answer 3

Rather than absolute quality, data quality is data being of sufficient quality for the purpose it’s being used for. This depends on how it’s used.

Question 4

How can errors be handled?

Answer 4

Accept
Reject
Correct (rectify)
Estimate (Interpolate/extrapolate)

Question 5

What are the Data Quality dimensions

Answer 5

AACCTUV

Accuracy
Appropriateness (correct data to solve problem)
Completeness (relevant data, not missing)
Consistency (compatible sources)
Timeliness (suitable temporal period)
Uniqueness (each record corresponds to a real world item)
Validity (data follows rules)

Question 6

What is the acronym relating to Data Security?

Answer 6

CIA
Confidentiality (auth access only)
Integrity (auth changes only)
Availability (when needed, may use SLA [Service Level Agreement])