Blocking Records Based on Triangle-Inequality

摘要

Deduplication and automatize the matching process is one of the most important research branch of database field. The paper proposes a novel method working in metrics space. Based on the triangle inequality theorem, the method blocks records into different, sets. Paper further proposes a multiple iterative mechanism and gives corresponding algorithms. The paper analyzes the complexity of the method and experimentally verifies the efficiency of the method. Compared with traditional methods, it improves the accuracy, precision and recall.