Fake news detection has been a tremendously challenging problem that affects real-world politics and information dissemination. The fact that content spreads so quickly and easily suggests that people (and algorithms behind the platforms) are potentially vulnerable to misinformation be it accidental or intentional. Despite the systematic efforts and fact-checking against misinformation, fake news still persists, which leads people to see and share information that is misleading.
Through this project our main objective would be to perform fake news detection using multiple models of varying complexity and provide a comprehensive comparative analysis of the same. We have undertaken models including Logistic Regression, Support Vector Machine, Convolutional Neural Network and Bidirectional-Long Short Term Memory. Above mentioned models were chosen in an attempt to reproduce the results of "Liar Liar pants on Fire"[3].
- LIAR Dataset (download link - [1])
- Kaggle Fakes News Dataset (download link - [2])
[1] LIAR Dataset: https://www.cs.ucsb.edu/~william/data/liar_dataset.zip
[2] Kaggle Fake News Dataset: https://www.kaggle.com/c/fake-news/data
[3] William Yang Wang, . ""Liar, Liar Pants on Fire": A New Benchmark Dataset for Fake News Detection." (2017).
[4] Krešňáková, Viera Maslej et al. "Deep learning methods for Fake News detection." 2019 IEEE 19th International Symposium on Computational Intelligence and Informatics and 7th IEEE International Conference on Recent Achievements in Mechatronics, Automation, Computer Sciences and Robotics (CINTI-MACRo).
[5] LibShortText: A Library for Short-text Classification and Analysis: https://www.csie.ntu.edu.tw/~cjlin/papers/libshorttext.pdf