Abstract: Multi-relational data mining is one of rapidly developing subfields of data minning. While classical data minning approaches look for patterns in a single relation, multi-relational data mining approaches look for patterns that involve multiple relations from a relational database. This paper provides a review of the research of multi-relational data mining. Firstly, the background and context that multi-relational data mining arises from is analyzed. Secondly, the general research methodologies of multi-relational data mining are summarized. Thirdly, typical algorithms of multi-relational data mining are introduced and analyzed. Finally, several challenging researching problems are identified.

Key words: Multi-relational data mining, Inductive Logic Programming, Multi-relational decision trees, Relational distance measure, Multi-relational association rules, Statistical Relational Learning

摘要： 多关系数据挖掘是近年来快速发展的重要的数据挖掘领域之一。传统的数据挖掘方法只能完成单一关系中的模式发现，多关系数据挖掘能够从复杂结构化数据中发现涉及多个关系的复杂模式。本文综述了多关系数据挖掘的研究状况。首先分析了多关系数据挖掘领域发生的原因和背景，其次总结了多关系数据挖掘研究的一般方法，再次介绍、分析了最具代表性的多关系数据挖掘算法。最后，总结了多关系数据挖掘将来发展需重点解决的问题和面临的挑战。

关键词: 多关系数据挖掘, 归纳逻辑程序设计, 多关系决策树, 关系距离测度, 多关系关联规则, 统计关系学习