欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
关联规则挖掘是数据挖掘领域的重要课题，但是，就评价关联规则是否有价值的依据，即兴趣度的度量方法，学术界没有一致的标准。传统的兴趣度度量方法包括支持度一置信度，提升度，改善度，有效度，影响度方法等。这些传统的兴趣度度量方法都存在各自的局限，本文首先比较分析了关联规则的客观兴趣度度量的相关研究成果，然后，针对它们的不足进行了改进，提出了两种比较有效的关联规则度量方法（New—lift，New—Improve），通过实验分析，进而提出新的度量框架，并实证了新方法的特征属性。
英文摘要：
The mining of association rules is an important topic issue in the domain of data mining. And there is no unitive interestingness measure method to decide whether the rule is valuable or not. Plentiful papers have discussed many interestingness measure methods such as the analysis of support-confidence, lift, improve, validity and influence. But these traditional measure methods have their own flaws. So this paper proposes two improved and more effective methods of interestingness measure based on the analysis on the comparison results of old methods, Then an improved measure framework is built through experimental analysis and the attributes of new methods and new framework are also proved.
同期刊论文项目
同项目期刊论文
Research on logistics network infrastructures based on dea-pca approach: Evidence from the yangtze r
Integrating the use of spreadsheet software in logistics education: An illustration of the use of Mi
A New Collaborative Recommendation Approach Based on Users Clustering Using Artificial Bee Colony Al
Personal Recommendation Using a Novel Collaborative Filtering Algorithm in Customer Relationship Man
期刊信息
位置：