DOI QR코드

DOI QR Code

CNN 기반 감성 변화 패턴을 이용한 가짜뉴스 탐지

Fake News Detection Using CNN-based Sentiment Change Patterns

  • 이태원 (한겨레신문(주) 독자서비스국 독자기획부) ;
  • 박지수 (전주대학교 컴퓨터공학과) ;
  • 손진곤 (한국방송통신대학교 컴퓨터과학과)
  • 투고 : 2023.01.31
  • 심사 : 2023.02.03
  • 발행 : 2023.04.30

초록

최근 가짜뉴스는 뉴스 콘텐츠 형식을 가장하고 중요한 사건이 발생할 때마다 등장하여 사회적 혼란을 초래한다. 이에 가짜뉴스를 탐지하기 위한 연구로 인공지능 기술이 사용된다. 자연어 처리를 통해 가짜뉴스를 자동으로 인지 및 차단하거나, 네트워크 인과 추론과 결합함으로써 허위 정보를 확산시키는 소셜미디어 인플루언스 계정을 감지하는 등의 가짜뉴스 탐지 접근법이 딥러닝을 통해 구현될 수 있었다. 그러나 가짜뉴스 탐지는 여러 자연어 처리 분야 중에서도 해결이 어려운 문제로 분류된다. 가짜뉴스가 가지는 형식 및 표현의 다양성으로 특성 추출의 난도가 높고, 뉴스가 속한 범주에 따라 하나의 특성이 서로 다른 의미를 가질 수도 있는 등 다양한 한계점이 존재한다. 본 논문에서는 가짜뉴스를 탐지하기 위한 추가적인 식별 기준으로 감성 변화 패턴을 제시한다. 합성곱 신경망을 가짜뉴스 데이터 세트에 적용하여 콘텐츠 특성에 기반한 분석을 수행하고, 감성 변화 패턴을 추가로 분석함으로써 성능이 개선된 모델을 제안한다. 뉴스를 구성하는 문장에 대하여 감성 극성을 산출하고 장단기 메모리를 적용함으로써 문장 순서에 의존적인 결괏값을 얻을 수 있다. 이를 감성 변화의 패턴으로 정의하고 뉴스의 콘텐츠 특성과 결합하여 가짜뉴스 탐지를 위한 제안 모델의 독립변수로 활용한다. 제안 모델과 비교 모델을 딥러닝으로 학습시키고 가짜뉴스 데이터 세트를 이용한 실험을 진행하여 감성 변화 패턴이 가짜뉴스 탐지 성능을 개선할 수 있음을 확인한다.

Recently, fake news disguises the form of news content and appears whenever important events occur, causing social confusion. Accordingly, artificial intelligence technology is used as a research to detect fake news. Fake news detection approaches such as automatically recognizing and blocking fake news through natural language processing or detecting social media influencer accounts that spread false information by combining with network causal inference could be implemented through deep learning. However, fake news detection is classified as a difficult problem to solve among many natural language processing fields. Due to the variety of forms and expressions of fake news, the difficulty of feature extraction is high, and there are various limitations, such as that one feature may have different meanings depending on the category to which the news belongs. In this paper, emotional change patterns are presented as an additional identification criterion for detecting fake news. We propose a model with improved performance by applying a convolutional neural network to a fake news data set to perform analysis based on content characteristics and additionally analyze emotional change patterns. Sentimental polarity is calculated for the sentences constituting the news and the result value dependent on the sentence order can be obtained by applying long-term and short-term memory. This is defined as a pattern of emotional change and combined with the content characteristics of news to be used as an independent variable in the proposed model for fake news detection. We train the proposed model and comparison model by deep learning and conduct an experiment using a fake news data set to confirm that emotion change patterns can improve fake news detection performance.

키워드

과제정보

이 논문은 한국방송통신대학교 학술연구비 지원을 받아 작성된 것임.

참고문헌

  1. David M. J. Lazer et al., "The science of fake news," Science, Vol.359, No.6380, pp.1094-1096, 2018. https://doi.org/10.1126/science.aao2998
  2. Y.-s. Hwang and O.-s. Kwon, "A study on the conceptualization and regulation of fake news: Focusing on self-regulation of internet service providers," Media and Law, Vol.16, No.1, pp.53-101, 2017. https://doi.org/10.26542/JML.2017.4.16.1.53
  3. N. Wingfield, M. Isaac, and K. Benner, "Google and Facebook take aim at fake news sites," nytimes.com. https://www.nytimes.com/2016/11/15/technology/google-will-ban-websites-that-host-fake-news-from-using-its-ad-service.html (accessed Nov. 30, 2021).
  4. T.-W. Yun and H.-C. Ahn, "Prediction of domestic fake news using text mining and machine learning," Journal of Information Technology Applications & Management, Vol.25, No.1, pp.19-32, 2018.
  5. D.-H. Lee, Y.-R. Kim, H.-J. Kim, S.-M. Park, and Y.-J. Yang, "Fake news detection using deep learning," Journal of Information Processing Systems, Vol.15, No.5, pp.1119- 1130, 2019. https://doi.org/10.3745/JIPS.04.0142
  6. Y. Hyun and N. Kim, "Text-based fake news detection methodology using news and social data," Journal of the Korean Society of Electronic Transactions, Vol.23, No.4, pp.19-39, 2018.
  7. J.-w. Cho, E.-b. Kim, H.-m. Kim, and M.-a. Son, "Development of attention network with added entity for fake news detection," Proceedings of the Fall Conference of the Korean Industrial Engineering Society, pp.2751-2755, 2019.
  8. Dale E. Brashers, "Communication and uncertainty management," Journal of Communication, Vol.51, No.3, pp.477-497, 2001. https://doi.org/10.1111/j.1460-2466.2001.tb02892.x
  9. C. Choi, "The effect of fact-checking efforts on fake news efforts: Based on uncertainty management theory," Communication Theory, Vol.17, No.1, pp.5-53, 2021. https://doi.org/10.20879/ct.2021.17.1.005
  10. F. Liu, A. Burton-Jones, and D. Xu, "Rumors on Social Media in disasters: Extending Transmission to Retrans-mission," Pacific Asia Conference on Information Systems, pp.49, 2014.
  11. H. Karimi, P. C. Roy, S. Saba-Sadiya, and J. Tang, "Multisource multi-class fake news detection," Proceedings of the 27th International Conference on Computational Linguistics, pp.1546-1557, 2018.
  12. Y. Yang et al., "TI-CNN: Convolutional neural networks for fake news detection," arXiv:1806.00749, 2018.
  13. Seoul National University Press Information Research Institute SNU Fact Check Center, "SNU Fact Check," [Internet], https://factcheck.snu.ac.kr (downloaded on November 30, 2021).
  14. S.-w. Oh and G.-h. Hwang, "An exploratory study on the requirements for the formal composition of facts: Suggestions through SNU FactCheck metadata analysis," Press Information Research, Vol.55, No.4, pp.54-98, 2018. https://doi.org/10.22174/jcr.2018.55.4.54
  15. Open source Korean language processor. Yoo Ho-hyun. Downloaded on 30 November 2021. [Internet], https://github.com/open-korean-text/open-korean-text
  16. Korean Sentence Splitter. Sang-Kil Park. Accessed: Nov. 30, 2021. [Internet], https://github.com/likejazz/koreansentence-splitter
  17. Wikipedia, The Free Encyclopedia and s.v. "Wikipedia: Database download," 2021. Distributed by Wikimedia. [Internet], https://dumps.wikimedia.org/kowiki/latest/kowiki-latest-pages-articles.xml.bz2
  18. Wikipedia, The Free Encyclopedia and s.v. "Convolutional neural network," [Internet], https://en.wikipedia.org/wiki/Convolutional_neural_network (accessed Nov. 30, 2021).
  19. Nitish Srivastava, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever and Ruslan Salakhutdinov, "Dropout: a simple way to prevent neural networks from overfitting," Journal of Machine Learning Research, Vol.15, No.1, pp.1929-1958, 2014.
  20. Sergey Ioffe and Christian Szegedy, "Batch normalization: Accelerating deep network training by reducing internal covariate shift," Proceedings of the 32nd International Conference on Machine Learning, Vol.37, pp.448-456, 2015.