摘要
文章对竞争对手识别的本质进行了分析,认为竞争对手识别是一个动态分类过程,指出该过程有3个难点。现有的竞争对手识别方法存在3个缺陷,因而无法有效地支持竞争对手识别过程。认为数据挖掘决策树算法可以应用于竞争对手识别,并详细描述了应用的过程,包括以下4个步骤:竞争对手属性选择、竞争对手数据准备、竞争对手数据预处理和建立竞争对手识别决策树。
This paper analyzes the essence of competitor recognition, argues that competitor recognition is a dynamic classifi- cation process, and points out there are 3 difficulties in the process. There are also 3 defects in the present competitor recognition methods, so they can not support the competitor recognition process effectively. The paper assumes that the decision tree algorithm for data mining is applicable to competitor recognition, and gives a detailed description of the application process, including the following 4 steps ： selection of competitor attribute, preparation of competitor data, pre-processing of competitor data, and estab- lishment of decision tree for competitor recognition.
出处
《情报理论与实践》
CSSCI
北大核心
2013年第3期1-5,24,共6页
Information Studies:Theory & Application
关键词
竞争对手
决策树
识别
模型
competitor
decision tree
recognition
model
