Korean Voice Phishing Detection Applying NER With Key Tags and Sentence-Level N-Gram
Seunguk Yu, Yejin Kwon, Minju Kim, Ki-Seong Lee
Abstract
Voice phishing is the criminal act of tricking others to transfer funds or to seek financial gain based on personal information obtained illegally. The importance of this crime is recognized worldwide, and technical solutions have been proposed to reduce the increasing damage. In this paper, we propose a process for Korean voice phishing detection by applying named entity recognition with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Key Tags</i> and Sentence-level N-gram. From the perspective of human, we collect financial counseling texts as non-phishing dataset since the victim confuses voice phishing with them. We select <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Key Tags</i> that are meaningful for distinguishing voice phishing and financial counseling texts and combine sentence bundles to effectively detect voice phishing. The experimental results, using ten types of machine learning models, were maintained when generalizing information by <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Key Tags</i> and improved when combining text bundles. We hope that the proposed process, which applies NER with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Key Tags</i> and sentence-level N-gram, can be effectively applied to other criminal scenarios in the future.