Rumour detection using graph neural network and oversampling in benchmark Twitter dataset

This repository contains the code for our paper "Rumour detection using graph neural network and oversampling in benchmark Twitter dataset".

Cite

@misc{patel2022rumourdetectionusinggraph,
      title={Rumour detection using graph neural network and oversampling in benchmark Twitter dataset}, 
      author={Shaswat Patel and Prince Bansal and Preeti Kaur},
      year={2022},
      eprint={2212.10080},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2212.10080}, 
}

@InProceedings{10.1007/978-981-16-2597-8_8,
author="Patel, Shaswat
and Shah, Binil
and Kaur, Preeti",
editor="Khanna, Ashish
and Gupta, Deepak
and Bhattacharyya, Siddhartha
and Hassanien, Aboul Ella
and Anand, Sameer
and Jaiswal, Ajay",
title="Leveraging User Comments in Tweets for Rumor Detection",
booktitle="International Conference on Innovative Computing and Communications",
year="2022",
publisher="Springer Singapore",
address="Singapore",
pages="87--99",
abstract="A novel technique is presented in this paper to detect rumors from tweets. Rumor is unverified information at the time of posting. To detect rumors in the tweets, we use transformer models BERT, RoBERTA, ALBERT, and DistilBERT. These techniques perform the feature extractor for input sequence consisting of source tweet and user comments on the source tweet. The key insight is that by understanding the context of the source tweet and the user comments the models can successfully classify the source tweet into rumor or non-rumor. This is based on the fact that users on social media sites try to classify any new information into rumor and non-rumor collectively by using comments. Our approach was able to produce better precision, recall, and F1 score over the state-of-the-art classifier that uses Conditional Random Fields (CRFs) to learn the context during the event.",
isbn="978-981-16-2597-8"
}