Search:
Match:
1 results

Analysis

This paper addresses the important problem of distinguishing between satire and fake news, which is crucial for combating misinformation. The study's focus on lightweight transformer models is practical, as it allows for deployment in resource-constrained environments. The comprehensive evaluation using multiple metrics and statistical tests provides a robust assessment of the models' performance. The findings highlight the effectiveness of lightweight models, offering valuable insights for real-world applications.
Reference

MiniLM achieved the highest accuracy (87.58%) and RoBERTa-base achieved the highest ROC-AUC (95.42%).