A fact extraction and verification framework for social media posts on Covid-19

Date

2023-01-01

Author

Temiz, Orkun
Taşkaya Temizel, Tuğba

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

104
views

0
downloads

Cite This

Social media has become popular for spreading and consuming information online. On the other hand, the high number of posts has increased the need for fact checking. In the COVID-19 pandemic, the lack of information on the disease paved the way for the spread of false information, negatively affecting public health and society. In this paper, a new zero-shot fact extraction and verification framework for informal user posts on COVID-19 against medical articles is proposed. The framework includes five main steps, which are pre-processing user posts, claim extraction, document & evidence extraction, and verdict assignment. The framework aims to classify user posts while presenting the related evidence set extracted from peer-reviewed medical articles about each claim in user posts, making it interpretable for end users. The proposed framework obtains on-par and stable performance compared with the state-of-the-art supervised techniques for classifying raw user posts (Coaid) and rumors collected from social media (COVID-19 Rumors Dataset). By utilizing the zero-shot capabilities of the present models in the literature, it achieves superior performance detecting newly emerged misinformation posts and topics.

Subject Keywords

COVID-19, Fake News, LATEX Fact Checking and Verification System, Misinformation Detection with Credible Information Retrieval, Natural Language Processing

URI

https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85161687818&origin=inward
https://hdl.handle.net/11511/104585

Collections

Graduate School of Informatics, Conference / Seminar