Article

Download article

Modified method for detecting fake news based on machine learning algorithms

Anatolii Sanzharovskyi, Vasyl Yurchyshyn

Received 12.12.2022, Revised 17.04.2023, Accepted 11.05.2023

Abstract

The object of research is the process of analyzing information in social media to identify fake news. The subject of research is the development of software of algorithmic-software method for detecting fake news. The aim of the work is to increase the average accuracy of the process of detecting fake news in social media by developing and implementing an algorithmic-software method for detecting fake news based on machine learning algorithms. Various methods of scientific research: analysis – to find out the advantages and disadvantages of existing methods for detecting fake news; comparison – when choosing the most optimal programming language and programming environment for developing software to detect fake news; a method of reviewing existing literature to detect fake news, including academic publications, technical reports, and online resources; peer review method, which obtained information on the effectiveness of various methods for detecting fake news have been used. Through the use of these methods, a comprehensive understanding of the problem of detecting fake news has been obtained and effective software for detecting fake news has been developed. The scientific novelty of the work lies in the fact that a modified algorithmic-software method for detecting fake news based on machine learning algorithms has been proposed, which differs from the existing methods by using an ensemble of three algorithms, the results of each of which are used to select more compact specialized models for subsequent algorithms, which ultimately allows to speed up the process of detecting fake news in the text by 30% compared to analogs, and reduce the average falsehood by 25%. The practical value of the results obtained in the work lies in the fact that the developed software of the algorithmic-software method for detecting fake news will help to reduce the spread of fakes and detect them

Keywords:

algorithmic-software method; machine learning algorithms; methods of fake detection and recognition; BERT; LSTM; Passive-Aggressive Classifier

https://doi.org/10.24025/2306-4412.2.2023.279984

Retrieved from Volume 28, No. 2, 2023

Pages 58-70

References Suggested citation

References

[1] A. Sanzharovsky, and V. Yurchyshyn, "Algorithmic-software method for detecting fake news based on machine learning algorithms", on Fifteenth Sci. Conf. of Undergraduates and Graduate Students Applied Mathematics and Computing (PMK-2022), Kyiv, Nov. 16-18, pp. 499-504, 2022 [in Ukrainian]. 

[2] Study: On Twitter, false news travels faster than true stories. [Online]. Available: https://news.mit.edu/2018/study-twitterfalse-news-travels-faster-true-stories-0308. Accessed on: Jan. 20, 2023. 

[3] Disinformation risk assessment: The online news market in the United States.  [Online]. Available: https://www.disinformationindex.org/country-studies/ 2022-12-16-disinformation-risk-assessmentthe-online-news-market-in-the-unitedstates/. Accessed on: Jan. 20, 2023.

[4] H. Alcott, and M. Gentzkow, "Social media and fake news in the 2016 election", Journal of Economic Perspectives, vol. 31 (2), pp. 211-236. doi: 10.1257/jep.31.2.211. 

[5] K. Shu, A. Sliva, S. Wang, J. Tang, and H. Liu, "Fake news detection on social media: A data mining perspective", ACM SIGKDD Explorations Newsletter, vol. 19, iss. 101, pp. 22-36, Sept. 2017. doi: 10.1145/3137597.3137600.

[6] K. Sharma, F. Qian, H. Jiang, and N. Ruchansky, "Combating fake news: A survey on identification and mitigation techniques", ACM Transactions on Intelligent Systems and Technology, vol. 10 (3), pp. 1-42, Apr. 2019. [Online]. Available: https://www.researchgate.net/publication/332434399_Combating_Fake_News_A_Survey_on_Identification_and_Mitigation_ Techniques. Accessed on: Jan. 20, 2023.doi: 10.1145/3305260.

[7] B. D. Horne, and S. Adali, "This just in: Fake news packs a lot in title, uses simpler, repetitive content in text body, more similar to satire than real news", ArXiv abs/1703.09398, 2017. [Online]. Available: https://www.semanticscholar.org/paper/This -Just-In%3A-Fake-News-Packs-a-Lot-inTitle%2C-Uses-Horne-Adali/f8366afaf58 bbb9db151a1168bb6f14b618955b4. Accessed on: Jan. 20, 2023.

[8] D. Rothman, "Transformers for natural language processing: Build innovative deep neural network architectures for NLP with Python, PyTorch, TensorFlow, BERT, RoBERTa, and more"; Birmingham, UK: Packt Publishing Ltd. Birmingham Mumbai, 2021.

[9] P. Bahad, P. Saxena, and R. Kamal, "Fakenews detection using bi-directional LSTMrecurrent neural network", ProcediaComput. Sci., vol. 165, pp. 74-82, 2019.[Online]. Available: https://doi.org/10.1016/j.procs.2020.01.072. Accessed on: Jan. 20,2023

[10] S. M. Padnekar, G. S. Kumar, and P. Deepak, "Bilstm-autoencoder architecture for stance prediction", in Proc. 2020 Int. Conf. on Data Science and Engineering (ICDSE), Kochi, India, pp. 1-5, Dec. 3-5, 2020.

[11] E. Amer, K.-S. Kwak, and S. El-Sappagh, "Context-based fake news detection model relying on deep learning models".  Electronics, vol. 11 (8), p. 1255, 2022. [Online]. Available: https://doi.org/10.3390/ electronics11081255. Accessed on: Jan. 20, 2023 

[12] A. Malakhov, A. Patruno, and S. Bocconi, "Fake news classification with BERT". [Online]. Available: http://ceur-ws.org/ Vol-2882/paper38.pdf. Accessed on: Jan. 20, 2023.

[13] D. Jacob, C. Ming-Wei, L. Kenton, and T. Kristina, "BERT: Pre-training of deep bidirectional transformers for language understanding". [Online]. Available: https://arxiv.org/pdf/1810.04805.pdf. Accessed on: Jan. 20, 2023.

[14] Fake News Detection Using PassiveAggressive Classifier. [Online]. Available: https://link.springer.com/chapter/10.1007/97 8-981-15-7345-3_13. Accessed on: Jan. 20, 2023.

[15] D. Arthur, and S. Vassilvitskii, "k-means++: The advantages of careful seeding". [Online]. Available: https://theory.stanford.edu/~sergei/ papers/kMeansPP-soda.pdf. Accessed on: Jan. 20, 2023..

[16] D. P. Kingma, and J. L. Ba, "Adam: A method for stochastic optimization". [Online]. Available: https://arxiv.org/abs/ 1412.6980. Accessed on: Jan. 20, 2023.

Suggested citation

Sanzharovskyi, A., & Yurchyshyn, V. (2023). Modified method for detecting fake news based on machine learning algorithms . Bulletin of Cherkasy State Technological University, 28(2), 58-70. https://doi.org/10.24025/2306-4412.2.2023.279984