Si-QR: QR Code Cyber Detection Application using Fine-Tune IndoBERT Algorithm for Digital Transaction Security

Rafila Anindya Ahmad

Raliyanti, M.Pd.

MAN 4 South Jakarta

Ciputat Raya Street No.5, DKI Jakarta

rafila.anindya.ahmad@gmail.com

LPB NASIONAL 2024

Introduction

The use of QR codes (Quick Response Codes) has become increasingly widespread due to their practicality in transactions and accessing information. However, surveys indicate that 54.3% of respondents consider QR codes at risk of misuse (Primary data, 2023). Fraud cases, such as counterfeit QRIS (QR Code Indonesian Standard) codes, have occurred in public areas in Jakarta. According to data from cnnindonesia.com (10/5/24), a "quishing" scam method revealed by the police has the potential to steal personal data, and identities, and cause financial losses. This research objective is to design a prototype application, Si-QR, based on machine learning using a novel algorithm that combines Fine-Tuned IndoBERT and TLV EMV QRCPS (EMVCo QR Code Payment Systems) to detect QR code authenticity and evaluate its accuracy rate....