Thai-English Name Matching Using Google Translation and Thai Romanization

Businesses or contracts with illegal persons or persons on sanctions lists are not tolerated by companies. Sanctions lists are used to combat international terrorism by depriving individuals, organizations and institutions associated with terrorist activities of funds and economic resources. Sanctions lists, which can be purchased or sourced, may not contain complete information. Some lists or parts of the lists contain only first and last names in English. Neither first and last names in Thai nor identification numbers are presented. In order to improve the possibility of finding Thai people in a sanctions list, the method of converting Thai first and last names to those in English together with searching a criterion was proposed. Thai transliteration systems, Google Translate (GT), and our proposed methods were studied to convert Thai first and last names, and the results were evaluated using Levenshtein distance, accuracy, and similarity. When applying the royin engine, thai2rom engine, GT and our proposed method, the average Levenshtein distances per first and last name were 6.273, 3.717, 2.308 and 2.157, respectively. To search an English sanctions list, the minimum distance between the thai2rom engine and the GT results can be used as a criterion. It was found that the average distance reduced to 1.981, showing more efficient method to find unfavourable persons.