量化扩展概念格的属性归纳及多粒度规则挖掘  

Research on attribute-oriented induction based on quantitative extended concept lattice and multi-granule association rule mining

作　　者：王德兴[1] 胡学钢[2] 刘晓平[2] 

机构地区：[1]上海海洋大学信息学院,上海201306 [2]合肥工业大学计算机科学与信息学院,安徽合肥230009

出　　处：《系统工程学报》2009年第1期54-61,共8页Journal of Systems Engineering

基　　金：国家自然科学基金资助项目(60673028);海外及港澳学者合作研究基金资助项目(60828005);上海市教育委员会科研创新基金资助项目(08YZ120);上海海洋大学博士科研启动基金资助项目(科07-289)

摘　　要：在知识发现过程中用户感兴趣的往往是一些高层次、适当概括的简化信息,面向属性的归纳是目前主要的数据归约方法,一般是仅考虑原始数据所提供简单的统计信息;而基于量化扩展概念格的属性归纳算法,既可进行AOI的单一属性归纳,也能进行多层、多属性的归纳,而且泛化的路径不是唯一的,在量化扩展概念格的哈斯图很容易找到合适的泛化路径和阈值,得到满足用户要求合理的属性归纳结果,同时可以多层、多维的不同粒度的关联规则,有助于不同粒度知识的聚焦,发现不同粒度知识之间的变换关系.High-level, more generational and reductive information is getting more interesting for users in KDD. Attribute-oriented Induction （AOI） has been already primarily used in data reduction, which gen-erally takes the statistical information from original data into account. However, attribute-oriented induction based on quantitative extended concept lattice not only performs the task of AOI, but also carries out the generalization with multi-level and multi-attribute, while the generalization paths are not one and only. Finally, the proper generalization paths and thresholds on the Hasse diagram of quantitative extended concept lattice can be found, the required reasonable results can be easily gotten, moreover, multi-granule association rules with multi-level and multi-attribute are mined, different granule knowledge can be easily focused, and the relationships of transforms between different granule knowledge are discovered rapidly.

关 键 词：AOI 概念格 概念层次 数据挖掘 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]