Rough set extension of Tcl for data mining

摘要

Rough set theory provides a collection of methods for extracting previously unknown data dependencies or rules from relational databases or decision tables. Various applications exist, and several different low-level libraries are available that provide different analysis algorithms. Unfortunately, in many real applications, it is not easy to incorporate these low-level libraries. To deal with this problem, in this paper we introduce an approach which adds a rough set extension to an existing scripting language, Tcl. The extended language, known as RSLTcl, provides an easy-to-use interface and has a great deal of flexibility. We use two newly developed Tcl commands to illustrate our approach. We also use an example to show how these commands can be used to induce rules from decision tables. Since Tcl commands can be easily accessed through web pages, our approach may have good potential for incremental and automated construction of knowledge-bases.