Skip to content

Data and Code for EACL 2023 Paper "Reading and Reasoning over Chart Images for Evidence-based Automated Fact-Checking"

License

Notifications You must be signed in to change notification settings

mubasharaak/ChartFC_chartBERT

Repository files navigation

Intro to repository

This repository contains code and data accompanying the paper "Reading and Reasoning over Chart Images for Evidence-based Automated Fact Checking." published at EACL 2023 (Findings).

Structure

The folder ./code contains all scripts to run the VL experiments as well as code for the ChartBERT model. The subfolder ./data/dataset_files contains the ChartFC dataset files as well as processed versions, e.g. for the different sequence generation approaches discussed in the paper.

Downloading data

The chart images can be downloaded over the following link:

https://emckclac-my.sharepoint.com/:u:/g/personal/k20116188_kcl_ac_uk/EUoa-BgJEwZKgkx5sWFfg6wBFrf1-hP83Xtr3PjoNrUm9Q?e=U3Idz2

Reference

Akhtar, Mubashara, Oana Cocarascu, and Elena Simperl. 2023. Reading and Reasoning over Chart Images for Evidence-based Automated Fact-Checking. arXiv preprint arXiv:2301.11843. https://arxiv.org/pdf/2301.11843 (accepted to EACL 2023).

...

About

Data and Code for EACL 2023 Paper "Reading and Reasoning over Chart Images for Evidence-based Automated Fact-Checking"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published