Abhinav Pandey*

Domain Expert and Author, Data Science, Machine Learning & Artificial Intelligence, India

*Corresponding Author: Abhinav Pandey, Domain Expert and Author, Data Science, Machine Learning & Artificial Intelligence, India.

Received: July 15, 2021; Published: September 07, 2021

Abstract

  One of the most heinous crimes of the present times, Human Trafficking has been increasing at an alarming rate globally affecting millions of men, women and children. Amongst the various types of human trafficking comprising forced employment, organ smuggling, child marriage, sex trafficking and debt bondage, the market for sex trafficking has been making the headlines worldwide. Traffickers exploit these men, women and children and force them into flesh trade sometimes as early as when a girl attains puberty. Sadly, there is truly little that any law enforcement agency could do to bring the notorious traffickers to justice as majority of the cases do not even get reported. The advent of Internet has added to the woes of law enforcement authorities as the Traffickers easily advertise online from the comfort of their homes anywhere in the world. Traffickers are easily able to dodge the authorities by continuously deploying innovative advertising patters like using non-standard English grammar, emojis, multiple victims advertised simultaneously etc. This makes it extremely difficult to filter human trafficking ads from the genuine online escort service ads. In this study, we propose a novel architecture which extends BERT to incorporate not just texts but also emojis, special characters and other advertisement language patterns for doing multi-class text classification of the online advertisements into varying possibilities of them being labelled as sex trafficking advertisements.

Keywords: Human Trafficking; Deep Learning; Natural Language Processing; Forced Labor; Emojis; Semi-Supervised Learning

Bibliography

1. Kangaspunta Kristinia., et al. “UNODC global report on trafficking in persons 2020”. (2021).
2. Devlin Jacob., et al. “BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding” (2021).
3. Tong Edmund., et al. “Combating Human Trafficking with Deep Multimodal Models”. ACL 2017 - 55th Annual Meeting of the Association for Computational Linguistics, Proceedings of the Conference (Long Papers), vol. 1, Association for Computational Linguistics (ACL), May (2017): 1547-1556.
4. Work in Freedom Reducing Vulnerability to Trafficking of Women and Girls in South Asia and the Middle East Supporting Informed Migration, Fair Recruitment and Decent Work (2018).
5. Mccarthy Lauren A. “Human Trafficking and the New Slavery”. Annual Review of Law and Social Science 10 (2014): 221-242.

Citation

Citation: Abhinav Pandey. “Text Classification for Human Trafficking Using Advanced Transformers". Acta Scientific Computer Sciences 3.10 (2021): 21-26.

Copyright

Copyright: © 2021 Abhinav Pandey. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.