Rumor detection using dual embeddings and text-based graph convolutional network

Abstract Social media platforms like Twitter and Facebook have gradually become vital for communication and information exchange. However, this often leads to the spread of unreliable or false information, such as harmful rumors. Currently, graph convolutional networks (GCNs), particularly TextGCN,...

Full description

Saved in:
Bibliographic Details
Main Authors: Barsha Pattanaik, Sourav Mandal, Rudra M. Tripathy, Arif Ahmed Sekh
Format: Article
Language:English
Published: Springer 2024-11-01
Series:Discover Artificial Intelligence
Subjects:
Online Access:https://doi.org/10.1007/s44163-024-00193-6
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849221068110168064
author Barsha Pattanaik
Sourav Mandal
Rudra M. Tripathy
Arif Ahmed Sekh
author_facet Barsha Pattanaik
Sourav Mandal
Rudra M. Tripathy
Arif Ahmed Sekh
author_sort Barsha Pattanaik
collection DOAJ
description Abstract Social media platforms like Twitter and Facebook have gradually become vital for communication and information exchange. However, this often leads to the spread of unreliable or false information, such as harmful rumors. Currently, graph convolutional networks (GCNs), particularly TextGCN, have shown promise in text classification tasks, including rumor detection. Their success is due to their ability to identify structural patterns in rumors and effectively use neighborhood information. We present a novel rumor detection model using TextGCN, which utilizes a word-document graph to represent rumor texts. This model uses dual embedding from two pre-trained transformer models: generative pre-trained transformers (GPT) and bidirectional encoder representations from transformers (BERT). These embeddings serve as node representations within the graph, enhancing rumor detection. Combining these deep neural networks effectively extracts significant contextual features from rumors. This graph undergoes convolution, and through graph-based learning, the model detects a rumor. We evaluated our model using publicly available rumor datasets, such as PHEME, Twitter15, and Twitter16. It achieved 88.64% accuracy on the PHEME dataset, surpassing similar models, and performed well on Twitter15 and Twitter16 with accuracies of 81.98% and 83.41%, respectively.
format Article
id doaj-art-492535560a3e4baf930d83f906f7cefc
institution Kabale University
issn 2731-0809
language English
publishDate 2024-11-01
publisher Springer
record_format Article
series Discover Artificial Intelligence
spelling doaj-art-492535560a3e4baf930d83f906f7cefc2024-11-24T12:35:33ZengSpringerDiscover Artificial Intelligence2731-08092024-11-014111510.1007/s44163-024-00193-6Rumor detection using dual embeddings and text-based graph convolutional networkBarsha Pattanaik0Sourav Mandal1Rudra M. Tripathy2Arif Ahmed Sekh3School of Computer Science and Engineering, XIM UniversitySchool of Computer Science and Engineering, XIM UniversitySchool of Computer Science and Engineering, XIM UniversityDepartment of Computer Science, UiT The Arctic UniversityAbstract Social media platforms like Twitter and Facebook have gradually become vital for communication and information exchange. However, this often leads to the spread of unreliable or false information, such as harmful rumors. Currently, graph convolutional networks (GCNs), particularly TextGCN, have shown promise in text classification tasks, including rumor detection. Their success is due to their ability to identify structural patterns in rumors and effectively use neighborhood information. We present a novel rumor detection model using TextGCN, which utilizes a word-document graph to represent rumor texts. This model uses dual embedding from two pre-trained transformer models: generative pre-trained transformers (GPT) and bidirectional encoder representations from transformers (BERT). These embeddings serve as node representations within the graph, enhancing rumor detection. Combining these deep neural networks effectively extracts significant contextual features from rumors. This graph undergoes convolution, and through graph-based learning, the model detects a rumor. We evaluated our model using publicly available rumor datasets, such as PHEME, Twitter15, and Twitter16. It achieved 88.64% accuracy on the PHEME dataset, surpassing similar models, and performed well on Twitter15 and Twitter16 with accuracies of 81.98% and 83.41%, respectively.https://doi.org/10.1007/s44163-024-00193-6Rumor detectionRumor classificationDual word embeddingGraph convolution network (GCN)TextGCN
spellingShingle Barsha Pattanaik
Sourav Mandal
Rudra M. Tripathy
Arif Ahmed Sekh
Rumor detection using dual embeddings and text-based graph convolutional network
Discover Artificial Intelligence
Rumor detection
Rumor classification
Dual word embedding
Graph convolution network (GCN)
TextGCN
title Rumor detection using dual embeddings and text-based graph convolutional network
title_full Rumor detection using dual embeddings and text-based graph convolutional network
title_fullStr Rumor detection using dual embeddings and text-based graph convolutional network
title_full_unstemmed Rumor detection using dual embeddings and text-based graph convolutional network
title_short Rumor detection using dual embeddings and text-based graph convolutional network
title_sort rumor detection using dual embeddings and text based graph convolutional network
topic Rumor detection
Rumor classification
Dual word embedding
Graph convolution network (GCN)
TextGCN
url https://doi.org/10.1007/s44163-024-00193-6
work_keys_str_mv AT barshapattanaik rumordetectionusingdualembeddingsandtextbasedgraphconvolutionalnetwork
AT souravmandal rumordetectionusingdualembeddingsandtextbasedgraphconvolutionalnetwork
AT rudramtripathy rumordetectionusingdualembeddingsandtextbasedgraphconvolutionalnetwork
AT arifahmedsekh rumordetectionusingdualembeddingsandtextbasedgraphconvolutionalnetwork