Data Management

Database Reverse Engineering Based on Association Rule Mining

Free registration required

Executive Summary

Maintaining a legacy database is a difficult task especially when system documentation is poor written or even missing. Database reverse engineering is an attempt to recover high-level conceptual design from the existing database instances. In this paper, the authors propose a technique to discover conceptual schema using the association mining technique. The discovered schema corresponds to the normalization at the third normal form, which is a common practice in many business organizations. The algorithm also includes the rule filtering heuristic to solve the problem of exponential growth of discovered rules inherited with the association mining technique.

  • Format: PDF
  • Size: 356.55 KB