中文摘要：

为解决数据集内关联数据自动创建问题，采取推导传递法，借助多阶知识发现的理论，提出一种数据集内关联数据的自动创建方法。具体来说，主要是借助Falcon这一关联数据搜索引擎对数据集内的各列进行基于RDF三元组的共现检索，达到表内各列关联发现的目的，进而解决数据集内部关联数据的自动创建问题，并进行了实证研究，通过对实验结果的分析，在一定程度上证明了本文提出的数据集内关联数据自动创建方法的可行性和科学性。

英文摘要：

To solve the problem of automatic construction of linked data in data set, this paper proposes an Automatic Construction Meth-od of Linked Data in Data Set based on co-word analysis. Specifically, two words are firstly selected from two columns of a data set. Then, these two words are searched in Falcon to see if they are co-occurrence in a RDF, and to find their relationships, which are used to construct linked data. Finally, an experiment is conducted and proves the feasibility of the method proposed.