Sunteți pe pagina 1din 21

Quality Data and Information

Why should we care?

By Dinah Mande

April 5,2012

Information vs. Data


Data raw measurements, facts Information organized, compiled Intelligence/Knowledge actionable

Should we care about the quality of data and information?

Cost billions of dollars each year


Absolutely! It is important!

Costs lives

Cases on Poor Data Quality


Sept 11, 2001 Terrorist Attacks Sept 30, 1999 NASA Orbiter Mishap May 6, 2010 Misprinted Stock Quote
Shoot down orders were untimely, reaching pilots after the planes had already hit the World Trade Center

$125 million orbiter lost because two different teams working on the project used different units of measurement (English and the conventional metric system)

Nasdaq misprinted Proctor & Gamble stock error in entering a b for billion instead of m for million

price at $39.37 a share. they traded at $56.00 a share. The misprint coupled with a traders

What is Data Analysis and why companies need do it?


It is a practice in which raw data is inspected Transformed Cleansed and organized so that useful information can be extracted from it Forms of Data Analysis Charts Graphs Textual writeups of data
.

Why? Because it can easily increase the revenue and help make the right marketing and strategic decisions

Types of Data
Quantitative data (1,2,3)

Categorical data
Qualitative data(Quality and Category)

Causes of poor data quality

Why Profile Your Data?


To avoid poor quality data companies need do data profiling. (examining the data such as database) Why?

Because it directly Why? Poor data quality is costly. It lowers affects the customer satisfaction, effectiveness and adds expense, and efficiency of business makes it more difficult to processes. run a business

1.Standardization

Clean & consistent data

Biggest enemy of data quality STREET Street St. St ST Str Street (DataFlux power studio has great features for handling standardization)

2. Major auto companies records


Spelling of the color Beige
Beage beige Bage Beige BEIGE

Query: How well did beige sell last month?

Good data is base for good information Information treated as products are intended to create value for an organization
Value = Benefit Costs Benefits not always money

Lets take a few examples Common Database Issues

Information Wear & Tear


When information is used acquired, moved, copied, transformed, augmented,

IT IS SUBJECT TO DAMAGE

When information is not used IT GROWS STALE

Information should be tracked, controlled, and managed as an asset

Information Product life Cycle (POSMAD vs. SDLC)

Comparison of Physical Product Manufacturing to Information Manufacturing

Product Manufacturing Input Process Output Raw Materials Assembly Line Physical Products

Information Manufacturing Raw Data Information System Information Products

References
http://en.wikipedia.org/wiki/Data_analysis ftp://ftp.boulder.ibm.com/software/uk/govern/data_quality_to_th e_enterprise.pdf

Questions & Answers

S-ar putea să vă placă și