Work

Ground Truth-Oriented Fake News Detection Using Sentiment-Aware Hybrid Ensemble Learning and Bi-Directional LSTM Networks

Year: 2026
Type: article
Abstract: Fake news is the spreading of misinformation or disinformation.The proliferation of fake news on social media makes it difficult for one to procure reliable information online.Our motive is for information to be freely and widely available to all, while maintaining the integrity of said information.Ground truth refers to the actual reality or a baseline truth.Existing research focuses on the importance of deciphering the ground truth value and into a feature-oriented approach [1].This paper contributes to the existing body of literature by meticulously selecting features and leveraging emotional cues through natural language processing (NLP), to address the issue of fake news detection.This is done by availing of the Truth Seeker [2] and Liar [3] datasets.The models used are an ensemble model which is a combination of Naive Bayes, passive aggressive classifier with a layer of deep neural network and Bi-directional LSTM (Long Short-Term Memory) networks with sentiment analysis.Our findings suggest that sentiment analysis can serve as a valuable tool for enhancing the detection of fake news.
Language: English
Open Access status: diamond
OpenAlex ID: W7225889149
FWCI:
Cites: 2
Cited by:
Related to: 10