Özet:

Abstract Fake news often aims to damage the reputation of a person or entity, or to generate personal gain. The lack of a scalable fake news classification strategy is particularly worrying. Since manually classifying fake news is a time-consuming task, automatic identification of fake news has attracted a lot of attention in the Natural Language Processing (NLP) community to help ease the activity of classifying fake news. In recent Indonesian language news dataset, existing machine learning algorithms such as KNN and Naïve Bayes are used in this task, however it suffers from the lack of the ability to capture the true (semantic) meaning of words; therefore, the context is slightly lost. To address limitations, this paper introduces a new prediction using ensemble transformer based deep learning pre-trained language model such as BERT, RoBERTa, and DistilBERT as features extraction method on social media data sources. Finally, the system takes the decision based on model averaging to make prediction. Our proposed work yields promising performance as it has outperformed similar existing works in the literature. More precisely, our results achieve a maximum accuracy of 0.887 and f1 measure score of 0.878 on the news dataset.

Anahtar Kelimeler: