Research of Attribute Reduction Algorithm about Knowledge System

Rough set (RS) theory is a mathematical tool to deal with vagueness and uncertainty effectively developed in recent years. It has been applied widely in data mining, artificial intelligence, decision support system, pattern recognition, etc. In RS theory, Attribute Reduction (AR) is one of the most important and key contents. Many scholars all over the world have made a considerable amount of research about it. This paper summarizes AR algorithms based on knowledge system, and lays stress on basic principles, existing problems and disadvantages about AR algorithms. The future direction and trends of AR research are discussed in the end.