An Improved Algorithm for Attribute Reduction Based on Database Technology

Attribute reduction is the core issue of Rough Set theory and it has become the focus and hot spot of research in the field of Rough Set. At present, many attribute reduction algorithms are based on main memory and need I/O of text to achieve, which largely limit the application of these algorithms. In order to make up for the drawback of traditional attribute reduction model, some scholars proposed that combining rough set theory with relational database and using SQL language to achieve the algorithm. On one hand it can improve the algorithm efficiency, on the other hand it can also facilitate the real database application. In this paper, we propose an improved reduction algorithm based on database technology which omittes seeking nuclear steps to improve the efficiency of the algorithm. We also optimize relation query so that it can apply to inconsistent decision-making table. Finally, an example is given to illustrate the effectiveness of the algorithm.