Multi-criteria Group Decision Making with Various Ordinal Assessments

摘要

This paper proposes a method for addressing multi-criteria group decision making (MCGDM) problems with various types of ordinal assessments. In addition to two existing types of ordinal assessments including ordinal scale and ordinal interval, a new type of ordinal assessment named as ordinal distribution is developed, which is profiled by a belief distribution on a ranking framework of alternatives. Ordinal scales and ordinal intervals are transformed into ordinal distributions to implement unification of three types of ordinal assessments. Internal consistency about transformation from ordinal scales and ordinal intervals into ordinal distributions is theoretically analyzed and proven. To find solutions, dominating degrees between two alternatives are constructed based on the unified ordinal distributions to determine possibility degree to which one alternative is superior to another. A control mode selection problem for an enterprise located in Changzhou, Jiangsu, China, who mainly manufactures core parts of high-speed train, is analyzed using the proposed method to demonstrate its applicability. Based on the data in the control mode selection problem, internal consistency about transformation from ordinal scales and ordinal intervals into ordinal distributions is verified. It is also verified by simulation in a general case.