基于实体间关系的数据空间实体解析技术

Entity resolution in dataspace based on the relationship

针对数据空间中大量异质数据没有统一的语义,无法进行基于属性值相似度的实体解析任务的问题,提出了从实体间关系进行实体解析的简单方法。通过决策结点和决策关系构建连接图,并通过连通分量算法进行冗余结点的删除和属性的继承。通过构建的小规模数据集进行了算法的验证。

Aiming at the problem that a large number of heterogeneous data in the dataspace do not have uniform semantics,entity resolution task based on the similarity of attribute values cannot be carried out.In this paper,a simple method for entity reso-lution from inter-entity relations is proposed to construct connection graph through decision nodes and decision relations,and de-lete redundant nodes and inherit attributes through connected component algorithm.In this paper,the algorithm is verified by con-structing a small-scale data set.