Analyses & Alternatives (분석과 대안)

Korea Consensus Institute(KCI) (사단법인 코리아컨센서스연구원)
권호 표지
DOI QR코드

DOI QR Code

Integration of AI, Causality, and Social Sciences: Understanding Social Phenomena through Causal Deep Learning

AI, 인과성, 사회과학의 통합: 인과 딥러닝을 통한 사회현상의 이해

  • Seog-Min Lee (Hanshin University)
  • 이석민 (한신대학교)
  • Received : 2024.09.20
  • Accepted : 2024.10.12
  • Published : 2024.10.31
Download PDF
⟨ Previous Next ⟩

Abstract

This paper explores the integration of artificial intelligence and causal inference in social science research, focusing on causal deep learning. We examine key theories including Pearl's Structural Causal Model, Rubin's Potential Outcomes Framework, and Schölkopf's Causal Representation Learning. Methodologies such as structural causal models with deep learning, counterfactual reasoning, and causal discovery algorithms are discussed. The paper presents applications in social media analysis, economic policy, public health, and education, demonstrating how causal deep learning enables nuanced understanding of complex social phenomena. Key challenges addressed include model complexity, causal identification, interpretability, and ethical considerations like fairness and privacy. Future research directions include developing new AI architectures, real-time causal inference, and multi-domain generalization. While limitations exist, causal deep learning shows significant potential for enhancing social science research and informing evidence-based policy-making, contributing to addressing complex social challenges globally.

이 연구는 사회과학 연구에서 인공지능과 인과추론의 통합, 특히 인과적 딥러닝에 초점을 맞추고, Pearl의 구조적 인과모델, Rubin의 잠재적 결과 프레임워크, Schölkopf의 인과적 표현 학습 등 주요 이론들을 검토하였다. 또한 딥러닝을 활용한 구조적 인과모델, 반사실적 추론, 인과 발견 알고리즘 등의 방법론을 논의하였다. 본 연구는 소셜 미디어 분석, 경제 정책, 공중 보건, 교육 분야에서의 응용 사례를 제시하며, 인과적 딥러닝이 복잡한 사회 현상에 대한 세밀한 이해를 가능케 함을 보여주고 있다. 또한 모델의 복잡성, 인과 식별, 해석 가능성, 그리고 프라이버시 같은 윤리적 고려사항 등 주요 과제들을 다루었다. 향후 연구 방향으로 새로운 AI 아키텍처 개발, 실시간 인과 추론, 다중 도메인 일반화 등을 제시하였다. 비록 한계점들이 존재하지만, 인과적 딥러닝은 사회과학 연구 강화와 증거기반 정책 수립에 상당한 잠재력을 보이며, 전 세계적인 복잡한 사회 문제 해결에 기여할 것으로 기대된다. 특히 본 연구는 빅데이터 환경에서의 인과관계 식별과 해석의 중요성을 강조하며, 전통적인 통계적 방법론과 최신 딥러닝 기술의 결합이 가져올 시너지 효과를 탐구하고 있다. 또한 이 분야의 발전이 사회과학 연구의 패러다임을 어떻게 변화시킬 수 있는지에 대한 논의를 제공함으로써, 향후 사회과학과 인공지능 기술의 융합 연구에 대한 방향성을 제시하고자 하였다.

Keywords

Acknowledgement

This work was supported by Hanshin University Research Grant

References

  1. Ahne, A., Khetan, V., Tannier, X., Rizvi, M. I. H., Czernichow, T., Orchard, F., ... & Fagherazzi, G. (2021). Identifying causal relations in tweets using deep learning: Use case on diabetes-related tweets from 2017-2021. arXiv preprint, arXiv:2111.01225.
  2. Alfakih, A., Xia, Z., Ali, B., Mamoon, S., & Lu, J. (2023). Deep causality variational autoencoder network for identifying the potential biomarkers of brain disorders. IEEE Transactions on Neural Systems and Rehabilitation Engineering. https://doi.org/10.1109/TNSRE.2023.3344995
  3. Athey, Susan, & Imbens, Guido W. (2015). Machine learning methods for estimating heterogeneous causal effects. arXiv preprint, arXiv:1607.06580.
  4. Bengio, Yoshua, Deleu, Tristan, Rahaman, Nasim, Ke, Ruijiang, Lachapelle, Samuel, Bilaniuk, Olexa, ... & Pal, Chris. (2019). A meta-transfer objective for learning to disentangle causal mechanisms. arXiv preprint, arXiv:1901.10912.
  5. Bhopale, A. P., & Tiwari, A. (2024). Transformer-based contextual text representation framework for intelligent information retrieval. Expert Systems with Applications, 238, 121629.
  6. Chen, Hanwen, Wu, Xia, & Taylor, Lucas. (2023). Meta-causal learning: Transferring causal knowledge across domains. Journal of Machine Learning Research, 24(115).
  7. Chernozhukov, Victor, Chetverikov, Denis, Demirer, Mehdi, Duflo, Esther, Hansen, Christian, Newey, Whitney, & Robins, Jamie. (2018). Double/debiased machine learning for treatment and structural parameters. The Econometrics Journal, 21(1), C1-C68.
  8. Deaton, Angus, & Cartwright, Nancy. (2018). Understanding and misunderstanding randomized controlled trials. Social Science & Medicine, 210.
  9. Dwork, Cynthia. (2006). Differential privacy. In Proceedings of the 33rd international conference on Automata, Languages and Programming.
  10. Finn, Chelsea, Abbeel, Pieter, & Levine, Sergey. (2017). Model-agnostic meta-learning for fast adaptation of deep networks. In International Conference on Machine Learning. PMLR.
  11. Ganin, Yaroslav, Ustinova, Evgeniya, Ajakan, Hana, Germain, Pascal, Larochelle, Hugo, Laviolette, Francois, ... & Lempitsky, Victor. (2016). Domain-adversarial training of neural networks. The Journal of Machine Learning Research, 17(1).
  12. Guo, Ruocheng, Cheng, Liang, Li, Jun, Hahn, Peter R., & Liu, Huan. (2020). A survey of learning causality with data: Problems and methods. ACM Computing Surveys, 53(4).
  13. Heckman, James J. (2005). The scientific model of causality. Sociological Methodology, 35(1).
  14. Johansson, Fredrik, Shalit, Uri, & Sontag, David. (2016). Learning representations for counterfactual inference. In International Conference on Machine Learning. PMLR.
  15. Knaus, Michael. (2020). Double machine learning based program evaluation under unconfoundedness. Econometrics: Mathematical Methods & Programming eJournal. https://doi.org/10.1093/ectj/utac015
  16. Lundberg, Scott M., & Lee, Su-In. (2017). A unified approach to interpreting model predictions. In Advances in Neural Information Processing Systems.
  17. Morgan, Stephen L., & Winship, Christopher. (2015). Counterfactuals and causal inference: Methods and principles for social research (2nd ed.). Cambridge University Press.
  18. Pearl, Judea. (2009). Causal inference in statistics: An overview. Statistics Surveys, 3.
  19. Pearl, Judea, & Mackenzie, Dana. (2018). The book of why: The new science of cause and effect. Basic Books.
  20. Peters, Jonas, Janzing, Dominik, & Scholkopf, Bernhard. (2017). Elements of causal inference: Foundations and learning algorithms. MIT Press.
  21. Ribeiro, Marco T., Singh, Sameer, & Guestrin, Carlos. (2016). "Why should I trust you?" Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.
  22. Rojas-Carulla, Marta, Scholkopf, Bernhard, Turner, Richard, & Peters, Jonas. (2018). Invariant models for causal transfer learning. The Journal of Machine Learning Research, 19(1).
  23. Rosenbaum, Paul R. (2002). Observational studies (2nd ed.). Springer.
  24. Rubin, Donald B. (1974). Estimating causal effects of treatments in randomized and nonrandomized studies. Journal of Educational Psychology, 66(5).
  25. Scholkopf, Bernhard, Locatello, Francesco, Bauer, Stefan, Ke, Nan R., Kalchbrenner, Nal, Goyal, Anirudh, & Bengio, Yoshua. (2021). Toward causal representation learning. Proceedings of the IEEE, 109(5).
  26. Schuler, Andreea, Baiocchi, Michael, Tibshirani, Robert, & Shah, Niladri. (2018). A comparison of methods for model selection when estimating individual treatment effects. arXiv preprint, arXiv:1804.05146.
  27. Shi, Chansoo, Blei, David, & Veitch, Victor. (2019). Adapting neural networks for the estimation of treatment effects. In Advances in Neural Information Processing Systems.
  28. Spirtes, Peter, Glymour, Clark, & Scheines, Richard. (2001). Causation, prediction, and search. MIT Press.
  29. Spirtes, Peter, & Zhang, Kun. (2016). Causal discovery and inference: Concepts and recent methodological advances. Applied Informatics, 3(1).
  30. Varian, Hal R. (2016). Causal inference in economics and marketing. Proceedings of the National Academy of Sciences, 113(27).
  31. Wang, Cheng, Chen, Jing, Xie, Zitao, & Zou, Ji. (2024, July). Research on education big data for student's academic performance analysis based on machine learning. In Proceedings of the 2024 Guangdong-Hong Kong-Macao Greater Bay Area International Conference on Education Digitalization and Computer Science.
  32. Wang, Liang, Adiga, Aniruddha, Chen, Jing, Sadilek, Adam, Venkatramanan, Srinivasan, & Marathe, Madhav. (2022). CausalGNN: Causal-based graph neural networks for spatio-temporal epidemic forecasting. AAAI, 12191-12199. https://doi.org/10.1609/aaai.v36i11.21479
  33. Zhang, Kun, Scholkopf, Bernhard, Muandet, Krikamol, & Wang, Zhifeng. (2013). Domain adaptation under target and conditional shift. In International Conference on Machine Learning. PMLR.