Record linkage using fuzzy sets for detecting suspicious financial transactions

Identifying suspicious financial transactions and linking relevant records is an important data related problem. An appropriate identification may improve fraud detection and international security. The main problems in this linking process are the missing data, errors in the entries or out of date entries. This study aims at developing an efficient method for identifying the suspicious transactions. This proposed methodology compares the information provided in the financial transactions with the black lists, and links similar entries. Different distance measures, including extensions of fuzzy sets can be used with the proposed method.