DSS Architecture

A DSS needs a two-layer architecture, according to professor Clyde Holsapple, who teaches DSS design at the University of Kentucky.

The DSS kernel is the problem processor. It accepts requests in nonprocedural languages and responds with text, graphics, multimedia or some form of output.

The problem processor draws upon the knowledge systems defined in the DSS. Those systems can offer procedural knowledge, such as an algorithm used to recognize patterns in data; descriptive knowledge, such as raw data itself; or reasoning knowledge, which feeds predefined inferences to the problem processor.

This architecture, Holsapple says, holds true for early DSS deployments and for so-called business intelligence (BI) applications today. Online analytical processing, BI and knowledge management "are all manifestations of decision-support systems," he says.

Copyright © 2002 IDG Communications, Inc.

7 inconvenient truths about the hybrid work trend
Shop Tech Products at Amazon