Using Data Mining to Build Integrated Discrete Event Simulations

Authors:
David A. Holland
Affiliations:
Computer Sciences Corporation, Arlington, United States 22201
Venue:
ICDM '08 Proceedings of the 8th industrial conference on Advances in Data Mining: Medical Applications, E-Commerce, Marketing, and Theoretical Aspects
Year:
2008

Citing 2
Cited 0

Requirements Engineering: Processes and Techniques

Requirements Engineering: Processes and Techniques
Subdue: compression-based frequent pattern discovery in graph data

Proceedings of the 1st international workshop on open source data mining: frequent pattern mining implementations

Quantified Score

Hi-index	0.00

Visualization

Abstract

Building a system from disparate software requires analysis to establish commonality of code. The ability of a data mining tool to extract repeating functional structures is the first step to reduce exploration, save development time, and re-use software components. This case study looks specifically at the application of graph-based data mining algorithms to code re-factoring. After writing a module to obtain a graph representation of a discrete event model, we built a tool around the University of Washington's SUBDUE package to find recurring patterns of logic. This resulted in cleaner code and increased awareness of code re-use.