Computer based Info Systems - Decision Support Systems

Data  | DBMS
Overview | 

Overview

Not all decisions are recurring in nature, some occur only once or infrequently and irregularly. Decision support systems assist managers who must make decisions that are not highly structured. The reports required cannot be predetermined as such a DSS needs greater flexibility than other information systems. the user must be able to request reports by defining their contents, similarly the data required may come from many forms.

Approaches to DSS vary from;

• direct access to real-time transactional data (operational data), through
• middleware to the extraction and downloading of operational data, to
• the PC for use with spreadsheets and end-user data analysis tools.

Problems include the negative impact on operational systems, complexity and diversity of data structures, inflexible end-user access, inconsistent data and overall lack of integration.

Information as an asset

Some scenarios that could be addressed by a Decision Support system - Rich Ghiossi - Tandem.

• Billionaire eccentric Ross Perot's neice goes into a bank and is refused a loan to buy a house. Perot takes his business elsewhere.
• An airline tracks a passengers flight schedule and sees they had a four hour delay on the last flight. The next time the passenger checks in they can say "I see Mr. X you had a flight delay last time, let us upgrdae your flight to first class"
• When you ring a bank or airline, you expect them to know the current situation and not have to go through everything all over again.

(Keenan,R 1996, Mar)

Categories

• Reporting

Production reporting tools

• Let companies generate regular operational reports or support high volume batch jobs, such as calculating and printing pay cheques.
• Tools include COBOL, specialised 4GL's.

Desktop report writers

• Inexpensive desktop tools designed for end-users
• In gereral have
  • Graphical user interface
  • Built in charting functions
  • Can pull data from a variety of sources
• Beginning to include object-oriented interfaces & online analytical processing (OLAP) analysis.
• Examples include Crystal reports.

• Managed query

  • Shield SQL from end-users by inserting a Meta software layer between teh database and user.
  • Meta is software that provides subject-oriented views of the database and supports point and click SQL creation.
  • Examples include Oracles Discoverer/2000, IBMs Decision Server.

• Executive information system (EIS)

Sophisticated software able to draw together data from the organisations databases in meaningful patterns. Uses A report showing simple measurements which act as an early warning signal of slipping standards.

• EIS consists of sophisticated software, that like MIS and DSS can draw together data from an organisation database into meaningful patterns.
• With an EIS it is designed so that with little spare time, a top executive can find it easy to use.
• Not only does EIS allow an executive to gain direct access about a company's performance, they also allow managers to communicate with other managers and share ideas and information.
• In some quarters EIS has become "enterprise information system" or even "enterprise intelligence system"

A full featured EIS system can comprise;

• an extraction tool for obtaining data from a database - sometimes called an OLAP (online analytical processor) server in decision support systems - which puts data in a convenient form for slicing, dicing and drilling.
• Database management system
• GUI front end capable of creating graphs, charts, tree diagrams, drillable maps, exception processing (variance from norm)

Terms

• Slicing and dicing - multidimensional data manipulation
• drill-down - a facility to take a progressively more detailed view.
• Critical success factors (CSF) - the idea that there are some simple measures which act as an early warning signal of slipping standards.
• Key performance indicator (KPI) - a measurable parameter used to measure standards.

Example products

• CDP Cognos tools such as Powerplay, Impromptu
• IRI Express database management front end tools
• Continuum Pilot's Lightship
• SAS SAS/EIS module
• Oracle's Express Analyser

Ref [127]Computerworld NZ May '95

• Online analytical processing (OLAP)

  • Provide an intuitive way to view corporate data.
  • The tools aggregate along common business subjects or dimensions then let users navigate through the hierarchies and dimension with the mouse.
  • Users can drill down, across or up levels or dimensions and pivot or swap out the data to change the data view.
  • Expected to be Java baased in the near future.
  • Web based tools the preferred delivery mechanism. Fundamental to the delivery process is a Web gateway that translates the Web Servers request (from the users Web browser) to the OLAP server.
  • Two strategies
    • OLAP - where queries are put into a data cache.
      Suited to financial analysis since data typically in small chunks.
    • ROLAP (Relational) - Analysis performed on the fly going directly to the database.
      Suited to marketing based applications such as point of sale software.
  • Example tools, Oracle's Express, Cognos Powerplay.

• Data mining

  • Use a variety of statistical algorithms to analyse correlation of data variables and look for interesting patterns.
  • Tools vendors include; Datamind (Datamind), SAS Institute.

( Eckerson, W.,1997 Apr 21)
[Rev: 17/3/98] © 1998 V/2-Com (Verhaart), P O Box 8415, Havelock North, New Zealand.