From Noise to Knowledge: Distill Social Media Opinions on Rumor Detection Exploiting LLMs
- 14 August 2026
- journal article
- research article
- Published by World Scientific Pub Co Pte Ltd in International Journal of Computational Intelligence and Applications
Abstract
The majority of social media research relies on datasets gathered from social media websites such as Reddit and Twitter. However, their intrinsic high noisy content reduces the performance of data-driven models. This restriction of noisy data has made data preparation techniques necessary. Current systems usually need to pay more attention to data quality and their performance heavily depends on noisy social media datasets. In this work, we concentrate on creating high-quality social media data for rumor detection tasks on the widely popular PHEME-9 dataset. Eliminating noisy responses is essential for misinformation tasks since it guarantees that the model has been trained on precise and pertinent data, improving its capacity to identify and validate rumors successfully. Large language models (LLMs) are used in this work to filter out irrelevant comments prior to the application of machine and deep learning techniques. This bifurcated approach helps to improve model accuracy and lower computing burden. We assume that this will further help in accurate rumor identification and can support environmental sustainability using less computational resources. Our proposed methodology shows an average improvement in the trained filtered models’ performance in terms of accuracy and F1-scores on six events in the PHEME-9 dataset. Further, to validate the effectiveness of the trained model, we performed interpretability and error analysis.Keywords
Funding Information
- EU H2020 program under the SoBigData++ project (871042)
- CHIST-ERA (CHIST-ERA-19-XAI-010)
- Eesti Teadusagentuur (SLTAT21096)
- Holistic Analysis of Organized Misinformation Activity in Social Networks (PCI2022-135026-2)
This publication has 49 references indexed in Scilit:
- Enhancing rumor detection with data augmentation and generative pre-trained transformerExpert Systems with Applications, 2024
- Rumour detection on benchmark twitter datasets using graph neural networks with data augmentationSocial Network Analysis and Mining, 2024
- Game-on: graph attention network based multimodal fusion for fake news detectionSocial Network Analysis and Mining, 2024
- A survey on rumor detection and prevention in social media using deep learningKnowledge and Information Systems, 2023
- SIRA: a model for propagation and rumor control with epidemic spreading and immunization for healthcare 5.0Soft Computing, 2022
- What goes on inside rumour and non-rumour tweets and their reactions: A psycholinguistic analysesComputers in Human Behavior, 2022
- Deep learning for misinformation detection on online social networks: a survey and new perspectivesSocial Network Analysis and Mining, 2020
- Fake news, rumor, information pollution in social media and web: A contemporary survey of state-of-the-arts, challenges and opportunitiesExpert Systems with Applications, 2020
- Microblog Rumor Detection Based on Comment Sentiment and CNN-LSTMLecture Notes in Electrical Engineering, 2020
- Content Representation for Microblog Rumor Detection2016