info-circle
student.uva.nl is vernieuwd. Lees meer...
Welke opleiding volg je?
Lezing

Presentation Master's Thesis - Max Toorians - Psychological Research Methods

Laatst gewijzigd op 01-08-2022
Automation of AML & Sanction Monitoring: Detection of Typographical Errors in Personal Names
Toon informatie specifiek voor jouw opleiding
Nu zie je algemene informatie op deze site. Kies je opleiding om ook informatie te zien die specifiek voor jouw opleiding geldt, zoals deadlines, regelingen en contactgegevens.
Welke opleiding volg je?
Startdatum
09-08-2022 09:00
Einddatum
09-08-2022 10:00
Locatie

Roeterseilandcampus - Building G

Street:

Nieuwe Achtergracht 129-B

Room: 

S.01

Banks and financial institutions have enormous costs in Anti Money Laundering (AML). A significant amount of costs is due to salary of compliance officers (CO) who are hired to prevent sanctioned transactions. In order to keep expenses sustainable, I propose an automation step in AML systems, or Transaction Filtering Systems. CO manually label potentially sanctioned transactions as true or false positive. Subsequently the hits are labeled with explanation by the CO. This thesis is an exploration, and first design of automatic labeling of personal names on typographical errors (typos). 

The thesis investigates two methods, an N-gram, and a Temporal Convolutional Network (TCN) approach to automate labeling of names with typos. Both approaches disentangle string structure, thus names are processed on a character level. The N-gram approach will train for di- and tri- grams within names and test on unseen names using a Maximum Likelihood Estimator (MLE). The TCN is a one-dimensional Convolutional Neural Network (CNN) that uses embedded tokens of characters to label names on typos. The societal relevance, methodological set up, results and discussion will be addressed in the thesis presentation.