Elba, Abdelrahman2024-06-112024-06-112024-04https://unbscholar.lib.unb.ca/handle/1882/37980As fake news spreads across languages, the challenge of detecting it in linguistically diverse environments becomes increasingly critical. Existing fake news detection methods are predominantly monolingual and heavily biased towards resource-rich languages. Fake news detection models often encounter difficulties with language-specific limitations and fail to account for linguistic and cultural variations in cross-lingual scenarios. This research contributes a comprehensive cross-language analysis, exploring the effectiveness of various detection models in both monolingual and cross-lingual contexts. This thesis emphasizes the importance of translation models and their role in enhancing detection accuracy. The attention-based model proposed significantly enhances the effectiveness of fake news detection. The comprehensive cross-language analysis demonstrates the model’s superiority over existing methods, it showcases a significant improvement in detection accuracy, with a 15% increase over conventional methods. The results show an improvement in accuracy, indicating a promising direction for future research.xii, 57electronicenhttp://purl.org/coar/access_right/c_abf2master thesisHakak, SaqibComputer Science