Record matching method based on local CON model in distributed database

Abstract

For existing record matching methods needed much artificial participation of experts or depend on heuristic rules heavily,and they could not handle the problem of large-scale data,A record matching method based on the local CON model was proposed.The approach used algorithm of association rules to get the match dependence,then took both match dependence and data instances as the input of the improved tableau,finally got the result by match detection.Ex-perimental results and theoretical analysis show that the method can quickly identify whether the distributed records matched,and it doesn’t require artificial participation,much better than previous methods.

Keywords

• distributed database;record matching;Apriori+;tableau