Authors: Hans W. A. Hanley, Deepak Kumar, Zakir Durumeric

Published on: August 03, 2023

Impact Score: 8.45

Arxiv code: Arxiv:2308.02068

Summary

• What is new: Introduction of an automated system using MPNet and DP-Means clustering to track news narratives on unreliable websites.
• Why this is important: Lack of automated tools for tracking misinformation across online platforms.
• What the research proposes: A system that identifies and tracks misinformation narratives using large-language models and clustering algorithms.
• Results: Identified 52,036 narratives across 1,334 websites, unveiling the most prevalent misinformation of 2022 and the key websites spreading them.

Technical Details

Technological frameworks used: nan

Models used: MPNet, DP-Means clustering

Data used: Daily scrapes of 1,334 unreliable news websites

Potential Impact

Fact-checking organizations, social media platforms, news agencies

Want to implement this idea in a business?

We have generated a startup concept here: NarrativeNet.