************************************************************
DSS News
D. J. Power, Editor
August 18, 2002 -- Vol. 3, No. 17
A Bi-Weekly Publication of DSSResources.COM
************************************************************
Check Larry English's article "Ten mistakes to avoid if
your data warehouse is to deliver quality information"
************************************************************
Featured:
* DSS Wisdom
* Ask Dan! - Is ETL software needed to build a model-driven DSS?
* Report from AMCIS 2002 - Dallas
* What's New at DSSResources.COM
* DSS News Releases
************************************************************
DSS News is sent to more than 850 subscribers from 50
countries. Please forward this newsletter to people interested
in Decision Support Systems and suggest they subscribe.
************************************************************
DSS Wisdom
According to Graham Allison, the Cuban Missile Crisis and technology
advances prompted a number of changes in military decision making.
Allison writes "Advances in the technology of communications made it
possible for political leaders in the White House to talk directly with
commanders of destroyers stationed along the quarantine line. Advances
in the technology of mass destruction created the possibility that acts
by men on a single destroyer in that quarantine line could rapidly
escalate to bring death to millions of Americans. Thus the governmental
leaders had both the capability and the incentive to reach out beyond
the traditional limits of their control. Maps in the 'Situation Room'
in the basement of the White House tracked the movement of all Soviet
ships. The members of the ExCom knew each of the ships by name and
argued extensively about which should be stopped first, at what point,
and how. Sorenson records 'the President's personal direction of the
quarantine's operation ... his determination not to let needless
incidents or reckless subordinates escalate so dangerous and delicate a
crisis beyond control.' Thus, for the first time in U.S. military
history, local commanders received repeated orders about the details of
their military operation directly from political leaders -- contrary to
two sacred military doctrines. This circumvention of the chain of
command and the accompanying countermand of the autonomy of local
commanders created enormous pain and surface friction. (p. 127-128)"
from Allison, G. T., Essence of Decision: Explaining the Cuban Missile
Crisis, Boston, MA: Little, Brown and Co, 1971.
************************************************************
Get Dan Power's new book, Decision Support
Systems: Concepts and Resources for Managers, at
http://greenwood.com/books/BookDetail.asp?dept_id=1&sku=Q497.
************************************************************
Ask Dan!
by Daniel J. Power
Is ETL software needed to build a model-driven DSS?
Extract, transform and load (ETL) tasks are part of building many
types of DSS, including some model-driven DSS. BUT, the ETL software
developed for creating and refreshing large data stores from
transaction, enterprise resources planning (ERP) and/or operating
systems is NOT needed for model-driven DSS.
In the Ask Dan! column titled "What is ETL software and how is it
related to DSS?" (DSS News 08/04/2002), the basics of this type of
software are discussed. The article by Larry English published at
DSSResources.COM on August 11, 2002 also has a discussion of ETL and
data quality. This column trys to clarify what data is needed for
model-driven DSS and how it is obtained.
Model-driven DSS use complex financial, simulation, and/or optimization
models to provide decision support. The needed data sets are usually
small, and certainly much smaller than the 500 megabyte-5 terabyte data
stores common with data-driven DSS.
In some model-driven DSS, the user enters ALL of the data needed by the
system. The DSS performs data validation and data storage. The data
entry may be 5-15 parameter values, text or other inputs. No data is
imported from a source system. For example, most of the model-driven
decision aids at DSSResources.COM have users input all of the data
required by the model. To try the Cost/Benefit Analysis decision aid
designed by D. J. Power and programmed in javascript by A. P. Power,
check URL http://dssresources.com/decisionaids/cbanalysis.html.
In other model-driven DSS, a time series of data on one or more
variables needs to be imported into the DSS. The data set may be 1000
to even 10,000 values. It is common to perform extract and transform
tasks to create the data. A report or data set is exported from a
source system. Then because the data set is small the data is usually
cleaned up and formatted in a text editor or in a desktop application
like Excel. Excel is often a useful tool for creating small data sets
for use with a model-driven DSS. The data set can then be incorporated
into a spreadsheet-based DSS built using Excel or Lotus 123 or imported
into another DSS development environment.
Larger data sets are used for some specialized model-driven DSS, but the
size of the data set remains modest compared to data marts and data
warehouses. For example, Radical Logistics sells transportation
software for calculating distances and rates. Data on "thousands of
shipments" is used for the analysis and the data needs to be verified
and cleaned up for correct ZIP codes and mileages. The Radical
Logistics software help with ETL, analysis and decision support.
Another common type of model-driven DSS uses a small number of data
values from an external database that is needed for the analysis by the
DSS user. The user defines the analysis and inputs some parameter
values. For example, many model-driven investment DSS extract data from
a historical stock market database. The Intrinsic Value per Share
Calculator at Quicken.com extracts earnings and price information from a
general purpose database of stock information and the user inputs
assumptions about interest rates for "What if?" analysis.
Finally, some model-driven DSS need very large data sets to create a
visual simualation that the DSS user can interact with. These data sets
are created and data may be imported from video files, maps and other
sources. For example, DaimlerChrysler has a Virtual Reality Center to
analyze and understand digital models and simulation results. The data
used is not ERP or transaction data. The extract, transform and load
tasks are very different from those associated with data warehousing,
business intelligence and data-driven DSS.
So ... As the data needs of a model or models in a model-driven DSS
increases, it becomes more likely that specialized software will be
needed to help the DSS developer create the specific decision support
data store. The software used to extract, transform and load the data
depends on the data, the DSS development environment and the preferences
of the developer.
References
English, L. P., "Ten mistakes to avoid if your data warehouse is to
deliver quality information", DSSResources.COM, 08/11/2002.
Power, D. J. "What is ETL software and how is it related to DSS?", DSS
News, 08/04/2002.
************************************************************
Report from AMCIS 2002 - Dallas
by Dan Power
Chair, AIS SIG DSS
AMCIS 2002 in Dallas had over 125 sessions. More than 300 papers were
presented and there were 17 panels and tutorials. The SIG DSS business
meeting was well attended and membership forms for SIG DSS should be
available in September. Contact SIG DSS secretary/treasurer Ramesh
Sharda, sharda@okstate.edu. So what were the highlights? Well ... the
sessions on decision support and data warehousing! Keith Lindsey and
Mark Frolick (Univ. of Memphis) discussed current issues in Data
Warehousing. Cliff Ragsdale (Virginia Tech) chaired a panel on
"Spreadsheet-based DSS Curriculum Issues". Cliff provided examples of
how he uses spreadsheets and Visual Basic for Applications to teach DSS
concepts. Paul Bergey (NC State) and I also discussed our teaching
approaches. A session on Web-based DSS identified some interesting
issues. In the final DSS paper session on Sunday morning, Steve Alter
and Roger Pick presented different views of DSS. Alter presented a work
system view of DSS that stimulated an interesting discussion and
comments. I heard positive comments on a number of other sessions, but
my highlights reflect only those sessions I could attend. Thanks to the
mini-track co-chairs for DSS - H. Bhargava, D. Power, M. Warkentin, and
for data warehousing - K. Dowling, D. Schuff, and R. St. Louis.
The Tex-Mex Fiesta on Saturday night was "big" even by Texas standards.
I'm sorry to report however, my opportunities to meet and chat weren't
as frequent as in Boston last year. Perhaps that will change next year.
Plans are in the works for AMCIS 2003, August 4-6, 2003 in Tampa, FL.
The conference theme is "Navigating the Torrents of Technology".
See you in Tampa.
************************************************************
What's New at DSSResources.COM
08/11/2002 Posted article by English, L. "Ten mistakes to avoid if your
data warehouse is to deliver quality information", DSSResources.COM,
08/11/2002, URL http://dssresources.com/papers/dssarticles.html.
************************************************************
DSS News Releases - August 5 to August 15, 2002
Complete news releases can be found at DSSResources.COM.
08/15/2002 A.D.A.M., Inc. and Subimo partner to create comprehensive
health care decision tools for consumers.
08/13/2002 Comshare(R) and Symmetry offer risk-free program to evaluate
analysis services.
08/12/2002 Radical Logistics has rapid adoption rate for new
transportation modeling solution.
08/08/2002 Intergraph's research program encourages innovation in the
geosciences.
08/08/2002 Roche Italy selects MicroStrategy Business Intelligence
Platform.
08/08/2002 Reminder: Americas Conference on Information Systems 2002 in
Dallas, TX August 9-11, 2002.
08/07/2002 Plumtree offers $75,000 portal package.
08/07/2002 Managers are unable to communicate value ... and it costs
their shareholders big-time.
08/06/2002 Hospitals reduce medical errors and improve care with
handheld technologies.
08/06/2002 New business planning products challenge investment doldrums.
08/05/2002 Reminder: The Data Warehousing Institute World Conference,
Las Vegas, NV, August 18-23, 2002.
08/05/2002 Primus Knowledge Solutions announces Primus® Quick Resolve -
A new product that accelerates the productivity of tier-one agents.
************************************************************
You can read 786 DSS News releases in
the Subscriber Zone at DSSResources.COM.
************************************************************
DSS News is copyrighted (c) 2002 by D. J. Power. Please send your questions to
daniel.power@dssresources.com.
|