---
title: "Can We Benefit from Metrics in KBS Development?"
authors:
  - "Stefan Kramer"
  - "Hermann Kaindl"
  - "Stefan Schlee"
doi: ""
url: "https://ijcai.org/proceedings/1997/97"
pdf: "https://ijcai.org/proceedings/1997/0097.pdf"
venue: "Proceedings of the Fifteenth International Joint Conference on Artificial Intelligence"
year: 1997
pages: "662-666"
track: "Expert Systems"
keywords:
---

# Can We Benefit from Metrics in KBS Development?

Stefan Kramer, Hermann Kaindl, Stefan Schlee

*Proceedings of the Fifteenth International Joint Conference on Artificial Intelligence*, pages 662-666.

## Abstract

A major issue with knowledge-based systems (KBS) is their maintenance. For dealing with this and other issues in software engineering for conventional software, there is a long tradition of using metrics. More recently, metrics have become popular also for object-oriented development. For th e development of a KBS , however, no m et rics are available yet. This pap er presents the co re of the first metrics suite for KBS development. We focus on the fundam e ntal concepts of coupling and cohesion, as we ll as on the impact of modifications. All these concepts are very important for ease of maintenance. The application of our metrics to a r ea l-world KBS based on rules and frames help ed pinpointing several maintenance proble ms . Th erefore, we propose to use our and other yet to be d Pve lop ed metrics in order to improve KBS deve lopment .