Abstract
The evaluation and decision taking methods proposed in this study bypass some final result obtained by performing traditional aggregation operators, and thus they provide some alternative way and choice for decision making and add diversity and flexibility as well. To achieve this purpose, three different types of partitions of input space are introduced, including a usual partition and two ordering determined partitions. Some relations and properties of those partitions are discussed. Several detailed construction methods for those different partitions are also proposed. A detailed decision problem is expressed by two different decision paradigms, with one relating to traditional aggregation operators and the other revolving around some decision functions which are based on partitions. As an extension, we also discuss some methods for generating new aggregation operators from given collections of aggregation operators.