摘要
在数据挖掘中SQL语句常用于数据的特征规则,特别是聚类规则。因为仅仅用一些简单的SQL语句就可以同时创建特征规则和聚类规则。把SQL语句和其他应用软件结合起来,将提高创建t-加权和d-加权的能力,特别是在特征规则中进一步归一化的能力,其中t-加权用于衡量每条数据记录的典型性,d-加权用于衡量聚类区分原则中的差异性。基于概念树把层次结构转换成表格结构将影响SQL语句的编写和成功执行,并且把层次结构中的每一个树结构转换为相应的表结构,最终把整个层次结构转换为表结构,才能合理高效地设计查询程序。
The SQL statement in data mining is usually used for the data feature rule, especially for clustering rule. With just few simple SQL statements, the characteristic and clustering rules can be created simultaneously. By combinating SQL statement with other application softwares, it can enhance the ability of creating t-weight for measuring the typicality of each data record in the characteristic rule and d-weight for measuring the difference in clustering discrimination rule, specifically for the ability of further normalization in characteristic rule. Converting hierarchy structure into table structure based on concept tree will influencethe compilation and execution of the SQL statement. The right method is to transform each of concept tree in concept hierarchy into a corresponding table structure first, and then transform the whole hierarchy structure into table structure. In this way, the query program can be designed reasonably and efficiently.
出处
《现代电子技术》
2012年第14期39-41,共3页
Modern Electronics Technique
关键词
数据挖掘
层次结构
聚类规则
SQL语句
data mining
hierarchy structure
clustering rule
SQL statement
