Deep Neural Networks in Social Media Forensics: Unveiling Suspicious Patterns and Advancing Investigations on Twitter

EasyChair Preprint 11083

Abstract

Text data forensics, an interesting field that focuses on careful analysis of textual content to identify criminal or suspicious activities, is developing at a rapid pace. With the popularity and the huge number of text posts on social media platforms, the demand for sophisticated forensic strategies has become critical. In this study, we enhance text data forensics in social media by leveraging the powerful analytical capabilities of deep neural networks. More specifically, we investigate the effectiveness of Long Short-Term Memory (LSTM) in the detection of suspicious text. The results are very promising as we achieved an accuracy of 96\% during preliminary evaluations. Many potential applications for the proposed model, including criminal activity identification, misinformation detection, and online harassment prevention that we plan to study in the future.

Keyphrases: Criminal activity, Deep Neural Networks (DNNs), Fake News., Natural Language Processing (NLP), Sentiment Analysis, Text data forensics, forensics investigations, online harassment, privacy and data protection, social media, social network analysis

@booklet{EasyChair:11083,
  author    = {Yousef Sharrab and Dimah Al-Fraihat and Mohammad Alsmirat},
  title     = {Deep Neural Networks in Social Media Forensics: Unveiling Suspicious Patterns and Advancing Investigations on Twitter},
  howpublished = {EasyChair Preprint 11083},
  year      = {EasyChair, 2023}}