Authors: Debarshi Brahma, Amartya Bhattacharya, Suraj Nagaje Mahadev, Anmol Asati, Vikas Verma, Soma Biswas

Published on: November 27, 2023

Impact Score: 7.8

Arxiv code: Arxiv:2311.16496

Summary

• What is new: Introduces a new framework, DPOD, which uses out-of-domain data to improve fake news detection in domains with limited data.
• Why this is important: The spread of fake news using out-of-context images across various domains with varying amounts of available data.
• What the research proposes: A novel framework, DPOD, which leverages out-of-domain data for improving multi-modal fake news detection.
• Results: Significantly surpasses existing approaches in detecting out-of-context misinformation on a large-scale benchmark dataset, NewsCLIPpings.

Technical Details

Technological frameworks used: DPOD (Domain-specific Prompt-tuning using Out-of-Domain data)

Models used: Modified Vision-Language Model, CLIP

Data used: NewsCLIPpings benchmark dataset

Potential Impact

Social media platforms, news agencies, digital content providers

Want to implement this idea in a business?

We have generated a startup concept here: TruthLens.