Development of a Multimodal System for Classifying and Automatically Reporting Illegal Online Prostitution Promotional Content Using Text and Image Data
Journal of Data Forensics Research · 31.12.2024 · 1 (1) · 29-46
DOI: 10.12972/jdfr.2024.1.1.3 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Data Forensics Research, online erschienen am 31.12.2024. Verfasst von Minhoi Park und 5 weiteren Autor:innen.
Abstract (Originalsprache)
Online prostitution advertisements and illegal content have emerged as serious social issues with the widespread use of the internet and social media platforms. Particularly, crimes targeting children and adolescents demand immediate and efficient responses due to their potentially far-reaching impacts. This study aims to develop an automated multimodal system for reporting sexual crimes, leveraging text and image data collected via the X API. The proposed system integrates a Multimodal Large Language Model (MLLM) to analyze text and image data comprehensively, enabling the detection of sexual crime indicators and the automatic generation of preliminary reports. These draft reports are reviewed by humans before being finalized and submitted, ensuring both speed and accuracy in the reporting process. Experimental results demonstrate that the developed system, by combining text and image data using MLLM, improves efficiency by over 30% compared to models trained on image data alone. This study contributes to the prevention of sexual crimes and the protection of victims, offering a practical tool and a novel approach for addressing illegal online content promoting prostitution.
Zitieren
APA
Park, M., Park, S., Yun, S., Choi, H., Kim, J., & Song, K. (2024). Development of a Multimodal System for Classifying and Automatically Reporting Illegal Online Prostitution Promotional Content Using Text and Image Data. Journal of Data Forensics Research, 1(1), 29-46. https://doi.org/10.12972/jdfr.2024.1.1.3
Vancouver
Park M, Park S, Yun S, Choi H, Kim J, Song K. Development of a Multimodal System for Classifying and Automatically Reporting Illegal Online Prostitution Promotional Content Using Text and Image Data. Journal of Data Forensics Research. 2024;1(1):29-46. doi:10.12972/jdfr.2024.1.1.3
BibTeX
@article{Park2024,
author = {Park, Minhoi and Park, Sumin and Yun, Sungbum and Choi, Hosik and Kim, Junchul and Song, Kyungwoo},
title = {Development of a Multimodal System for Classifying and Automatically Reporting Illegal Online Prostitution Promotional Content Using Text and Image Data},
journal = {Journal of Data Forensics Research},
year = {2024},
volume = {1},
number = {1},
pages = {29--46},
doi = {10.12972/jdfr.2024.1.1.3}
}
EUVIMED