中文摘要：

粒计算是知识表示与数据挖掘的一种重要方法.它模拟人类思考模式,以粒为基本计算单位,以处理大规模复杂数据和信息等建立有效的计算模型为目标.在粒计算看来,一个粒是由多个比较小的颗粒组成更大的一个单元.在许多场合下,由于不同标记尺度对数据有不同的分割,会得到不同层次的信息粒度.在面对具体问题时,自然而然地考虑选择一个合适的粒度层次来解决问题.针对具有多层粒度决策系统的粒度选择与规则获取问题,首先介绍了多粒度决策系统的概念,并在多粒度决策系统中定义了局部最优粒度,然后介绍了多粒度决策系统中基于局部最优粒度的属性约简.最后,给出了基于局部最优粒度的规则获取方法,并结合具体实例给出了规则获取的一个算法.

英文摘要：

Granular computing is an important issue in knowledge representation and data mining.It imitates thinking mode of human beings and its objective is to establish effective computation models and to seek for an approximation scheme for dealing with large scale complex data and information.In many real-life applications,it is impossible or unnecessary to distinguish individual objects or elements in the universe of discourses.With the view point of granular computing,the notion of a granule may be interpreted as one of the numerous small particles forming a larger unit.In many situations,there are different granules at different levels of scale in data sets having hierarchical scale structures.Many people focus on granular computing for problem solving and information processing by considering multiple levels of granularity.They all consider different descriptions of the same problem at multiple levels of granularity.This allows us to focus on solving aproblem at the most appropriate level of granularity by ignoring unimportant and irrelevant details.Due to the rampant existence of multi-granular decision systems in real world,the purpose of this study is to select appropriate level of granularity and to discuss rule acquisition from multi-granular deci-sion systems.The concept of multi-granular decision systems is introduced firstly.The notions of the optimal granularity and the local optimal granularity in a multi-granular decision system are then defined.Attribute reductions based on the local optimal granularities are further explored.Finally,rule acquisition in a multi-granular decision system is also discussed,and an algorithm of rule acquisition is illustrated with an example.