摘要：

传统挖掘方法容易遗漏由频率高的项目集构成的规则，且其作用度取值范围不够规范。针对该问题，提出一种基于对比影响度的正负关联挖掘方法，可使量化结果控制在[-1, 1]间，有效挖掘正负关联规则。将该方法应用到安徽巢湖地区性患病率高的食道癌病例数据关联规则的知识发现中，能挖掘出患食道癌的相关潜在因素，并给出对应预防措施。

关键词: 正负关联规则, 对比影响度, 食道癌

Abstract:

Traditional mining method is easy to omit rules which is comprised by high frequent item set and it is non-standard for function degree value range. Aiming at these problem, this paper presents a positive and negative association rule mining method based on comparison influence degree. It can control the quantized result between [-1, 1], effectively mine positive and negative association rule. Application of this method into esophagus cancer case data which is a high incidence in Chaohu of Anhui, it can mine some potential relevant factors and give some corresponding measures to guard against esophagus cancer.

Key words: positive and negative association rule, comparison influence degree, esophagus cancer

中图分类号: 

• TP311