Protecting Girls from Harassment and Fraudulent Calls: A Voice-to-Text Approach

Abstract

The rise in harassment calls and fraud, particularly targeting girls, has resulted in adverse consequences including psychological distress and, in extreme cases, suicide. Furthermore, fraudulent calls urging individuals to click on malicious links have led to substantial financial losses. This study presents a comprehensive approach to address this challenge for the development of an innovative detection system. Additionally, we introduce a novel prototype that employs a voice-to-text approach to transcribe phone calls, utilizing Natural Language Processing (NLP) techniques as well as Machine Learning (ML) algorithms to identify harassment or fraud-related content. When a malicious call is detected, the system automatically alerts parents or guardians and the nearest police station to prevent tragic outcomes such as suicides among targeted girls and financial fraud. By focusing on both preventive measures and advanced detection, this integrated approach aims to promote a safer communication environment and a more inclusive society.