Abstract:
Concerning telecom data mining project implementation and the data requirement of one business question, this paper analyzes the data preparation process, proposes a canonical plan for these items: wide-table structure design, source system data interface, data pretreatment process. It also pre-defines the of data exploration & data preparation to improve the of data mining project implementation efficiency and quality.
Key words:
data mining,
data preparation,
wide-table,
canonical
摘要： 从工程化实施电信数据挖掘项目的角度出发，在满足具体商业问题建模的数据要求前提下，对数据准备过程进行了结构化的分析和分解，提出一种规范化方法来约束宽表结构、源系统接口方式、数据预处理流程，并且预定义了相应的数据探索和数据准备过程，从源头改进电信数据挖掘项目的实施效率和质量。
关键词:
数据挖掘,
数据准备,
宽表,
规范化
CLC Number:
CAI Xin. Data Preparation Process Canonical Design for Telecom Data Mining[J]. Computer Engineering, 2007, 33(24): 44-45.
蔡 鑫. 电信数据挖掘数据准备过程的规范化设计[J]. 计算机工程, 2007, 33(24): 44-45.