Optimization of edit distance algorithm for sanctions screening risk score assessment

Abstract

Evaluating Risk Score Assessment for sanctions screening is necessary to calculate and gauge the risk rate of data elements involved during screening. It includes string matching process in reviewing sanctions lists to check if any investor in a fund is involved in fraud by matching the investor information (as Stop Descriptor) with the Sanctions List which contains the names of individuals who are known to be involved in financial crime or terrorism. This paper will present the inherent capability of Edit Distance algorithm or the Damerau-Levenshtein (DL) Distance algorithm to address many common misspellings and typos in string matching through insertion, deletion, transposition and substitution which are considered as a significant component of fuzzy possible success rating used in Sanction Screening. The paper also aims to optimize the DL Distance Algorithm by applying the theories of phonetic algorithm which expected to provide big impact on speed performance problem of computing the edit distance of two longer strings.