Enhancing misinformation detection using long short-term memory (LSTM) and bidirectional LSTM (Bi-LSTM) with word embedding techniques

Abstract

Misinformation is a pervasive issue in today’s society, with the spread of false or misleading information having potentially far-reaching consequences. In recent years, there has been a growing interest in using Artificial Intelligence (AI) technologies, such as Natural Language Processing (NLP) and machine learning, to detect and combat the spread of misinformation. In this study, we compare the performance of Long Short-Term Memory (LSTM) and Bidirectional LSTM (Bi-LSTM) in misinformation detection. We conducted experiments on two public datasets: “ISOT misinformation Dataset”. We trained LSTM and Bi-LSTM models on the preprocessed datasets and evaluated their performance using various evaluation metrics such as accuracy, precision, recall, and F1-score.