Authors: Tanveer Khan, Fahad Sohrab, Antonis Michalas, Moncef Gabbouj

Published on: February 03, 2024

Impact Score: 8.2

Arxiv code: Arxiv:2402.02066

Summary

• What is new: Introduction of a subspace-learning-based approach with a novel regularization term for One-Class Classification (OCC) models in classifying $\mathbb{X}$ (Twitter) users.
• Why this is important: Assigning credibility to $\mathbb{X}$ users and differentiating trusted from untrusted users using traditional machine learning methods has limitations.
• What the research proposes: Implementing a subspace-learning-based approach that optimizes subspace and data description for OCC, enhanced with a novel regularization term for Subspace Support Vector Data Description (SSVDD).
• Results: The introduced regularization term for SSVDD shows superior performance in classifying $\mathbb{X}$ users, outperforming baseline models and state-of-the-art techniques.

Technical Details

Technological frameworks used: Subspace Support Vector Data Description (SSVDD) with a novel regularization term

Models used: One-Class Classification (OCC) models

Data used: $\mathbb{X}$ user data

Potential Impact

Social media platforms for enhancing user trust and credibility assessment, cybersecurity companies focusing on social media protection

Want to implement this idea in a business?

We have generated a startup concept here: TrustGraph.