CityU launches new project to analyze online fraud tactics using AI
The total loss due to fraud worldwide has reached US$1 trillion, with losses in the Asia region accounting for 67% of the global total, amounting to HK$5.4 trillion.
In Hong Kong, as of October this year, the police recorded over 36,000 fraud cases, an increase of 7.3% compared to the same period last year, with the total loss amounting to HK$7.04 billion.
To better address the harm caused by fraud to society and the economy, a team from City University of Hong Kong has collaborated with China Unicom on a big data and artificial intelligence (AI) project, focusing on global telecom and online fraud issues, particularly the impact of telecom fraud on society and the economy. The project analyzes the techniques used by fraudsters, reveals their strategies for manipulating victims' trust and emotions, and delves into the human factors behind fraud.
The project was recently selected as one of the best practices for building a community of shared future in cyberspace at the World Internet Conference 2024.
The project integrates multi-source data from various regions worldwide, including police reports, media coverage, and social media, to establish a global fraud case database. It employs big data and AI technologies for in-depth analysis of fraud cases. At the same time, using large language models, it accurately depicts the behavioral patterns, rhetorical strategies of fraudsters, and the psychological characteristics of victims, thereby enhancing fraud prevention capabilities and public awareness.
The team has also closely collaborated with China Unicom to send over 1.5 billion public service text messages, conduct nearly 4,000 anti-fraud promotional activities, and issue 30 million precise warnings, thereby raising societal awareness of fraud prevention. They have improved the accuracy of the fraud call filtering system by 40% and reduced the false positive rate by 50%, gaining widespread recognition globally.
Related News:
HK fraudsters impersonating PayMe employees sentenced to imprisonment for 24 to 56 months
Comment