June 14, 2022. A team of researchers from Faculty of Management Sciences, Nakorn Ratchasima Rajabhat University, is developing the surveillance and monitoring system for human trafficking content on social media run by AI.
With the more obscure and complicated nature of child prostitution cases done through different channels, such as social media or game shops, more and more children are vulnerable, especially homeless children, child migrants, minority groups, refugees and stateless children. Study found a number of children engaged in prostitution voluntarily, most are between 15-17 years old.
The AI tool is developed to detect social media and pornography contents that may lead to child prostitution, both voluntary or forced. The system is built to be measurable and evaluable and is expected to help reduce budget and human resources. With the system that can screen and alert, it is hoped to lead to the decrease in child prostitution and trafficking cases and eventually contribute to the upgrade of Thailand’s ranking on TIP report.
This particular study started with data collection from Twitter posts and hashtags with relevance and relation to human trafficking and the interview with 30 anti-human trafficking police officers. The collection of keywords from the interview and 261,022 tweets was verified and categorized and then included in the Thailand Trafficking Corpus. The Corpus will in turn help increase the accuracy and efficiency of social media information searching.
The system prototyping model used 66,960 tweets to train with up to 89% accuracy and 16,740 tweets to test the performance.
