Question

Chapter 10 • Data Quality and Integration 

10-31

your instructor to download from the book's Web site. Your 
instructor can make those databases available to you. 
Alternatively, these and other databases are available at 
www.teradatastudentnetwork.com (your instructor will 
tell you the login password, and you will need to register 
and then create an SQL Assistant log-in for the parts of this 
question). There may actually be another database your 
instructor wants you to use for this series of questions. 
Regardless of how you gain access to a database, answer 
the following exercises for that database. 

a. Develop a plan for performing a data profile analysis on 
this database. Base your plan on the eight characteristics 
of quality data, on other concepts introduced in the 
chapter, and on a set of business rules you will need to 
create for this database. Justify your plan. 

b. Perform your data profile plan for one of the tables in 
the database (pick the table you think might be the most 
vulnerable to data quality issues). Develop an audit 
report on the quality of data in this table. 

c. Execute your data profile plan for a set of three or four 
related tables. Develop an audit report on the quality of 
data in these tables. 

d. Based on the potential errors you discover in the data to 
the previous two exercises (assuming that you find 
some potential errors), recommend some ways the cap- 
ture of the erroneous data could be improved to pre- 
vent errors in future data entry for this type of data. 

e. Evaluate the ERD for the database. (You may have to 
reverse-engineer the ERD if one is not available with the 
database.) Is this a high-quality data model? If not, how 
should it be changed to make it a high-quality data 
model? 

f. Assume that you are working with a Pine Valley 
Furniture Company (PVFC) database in this exercise. 
Consider the large and small PVFC databases as two 
different source systems within PVFC. What type 
of approach would you recommend (consolidation, 
federation, propagation, master data management), 
and why, for data integration across these two data- 
bases? Presume that you do not know a specific list of 
queries or reports that need the integrated database; 
therefore, design your data integration approach to 
support any requirements against any data from these 
databases.