A Novel Rough Set-based Framework for Objects Ranking in Ordered Information System

摘要

Aiming at improving the efficiency of the objects ranking in ordered information system, the problem of sorting is researched in this paper. Then a novel framework based on rough set theory is proposed the key idea of which is compressing the original data by attribute and value reduction. In the new approach the information table is simplified firstly by rough set theory after transforming an ordered information system into a binary one. On this basis the attribute values in the transformed information are standarded. And then a weight function is structured to evaluate the importance of each objects. Thus the objects can be sorted according to the value of weigh function. The experiment shows that the framework not only needn't extra information and the participation of the domain experts in ordering objects but also is feasible and effective.