Journal of The Korean Association For Science Education (한국과학교육학회지)

The Korean Association for Science Education (한국과학교육학회)
권호 표지
DOI QR코드

DOI QR Code

Understanding of Generative Artificial Intelligence Based on Textual Data and Discussion for Its Application in Science Education

텍스트 기반 생성형 인공지능의 이해와 과학교육에서의 활용에 대한 논의

  • Received : 2023.05.28
  • Accepted : 2023.06.07
  • Published : 2023.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

This study aims to explain the key concepts and principles of text-based generative artificial intelligence (AI) that has been receiving increasing interest and utilization, focusing on its application in science education. It also highlights the potential and limitations of utilizing generative AI in science education, providing insights for its implementation and research aspects. Recent advancements in generative AI, predominantly based on transformer models consisting of encoders and decoders, have shown remarkable progress through optimization of reinforcement learning and reward models using human feedback, as well as understanding context. Particularly, it can perform various functions such as writing, summarizing, keyword extraction, evaluation, and feedback based on the ability to understand various user questions and intents. It also offers practical utility in diagnosing learners and structuring educational content based on provided examples by educators. However, it is necessary to examine the concerns regarding the limitations of generative AI, including the potential for conveying inaccurate facts or knowledge, bias resulting from overconfidence, and uncertainties regarding its impact on user attitudes or emotions. Moreover, the responses provided by generative AI are probabilistic based on response data from many individuals, which raises concerns about limiting insightful and innovative thinking that may offer different perspectives or ideas. In light of these considerations, this study provides practical suggestions for the positive utilization of AI in science education.

본 연구는 최근 주목받고 있는 텍스트 기반 생성형 인공지능에 대해 관심과 활용이 증가함에 따라 과학교육적 측면에서의 활용을 위해 생성형 인공지능의 주요 개념과 원리를 설명하고, 이를 효과적으로 활용할 수 있는 방안과 그 한계를 지적하며 이를 토대로 과학교육의 실행과 연구의 측면에서 시사점을 제공하는 것을 목적으로 한다. 최근 들어 증가하고 있는 생성형 인공지능은 대체로 인코더와 디코더로 이뤄진 트랜스포머 모델을 기반으로 하고 있으며, 인간의 피드백을 활용한 강화학습과 보상 모델에 대한 최적화, 문맥에 대한 이해 등을 통해 놀라운 발전을 이루고 있다. 특히, 다양한 사용자의 질문이나 의도를 이해하는 능력과 이를 바탕으로 한 글쓰기, 요약, 제시어 추출, 평가와 피드백 등 다양한 기능을 수행할 수 있다. 또한 교수자가 제시하는 예를 토대로 주어진 응답을 평가하거나 질문과 적절한 답변을 생성하는 등 학습자에 대한 진단과 실질적 교육내용의 구성 등 많은 유용성을 가지고 있다. 그러나 생성형 인공지능이 가지고 있는 한계로 인해 정확한 사실이나 지식에 대한 잘못된 전달, 과도한 확신으로 인한 편향, 사용자의 태도나 감정 등에 미칠 영향의 불확실성 등에 대한 문제 등에 대해 해가 없는지 검토가 필요하다. 특히, 생성형 인공지능이 제공하는 응답은 많은 사람들의 응답 데이터를 기반으로 한 확률적 접근이므로 매우 거리가 멀거나 새로운 관점을 제시하는 통찰적 사고나 혁신적 사고를 제한할 우려도 있다. 이에 따라 본 연구는 과학교수학습을 위해 인공지능의 긍정적 활용을 위한 여러 실천적 제언을 제시하였다.

Keywords

Acknowledgement

이 연구는 2023학년도 단국대학교 대학연구비 지원으로 연구되었음.

References

  1. Bailey, J. O., Patel, B., & Gurari, D. (2021). A perspective on building ethical datasets for children's conversational agents. Frontiers in Artificial Intelligence, 4, 637532. 
  2. Boden, M. A. (2018). Artificial intelligence: A very short introduction. Oxford: Oxford University Press. 
  3. Cameron, L. (2002). Metaphors in the learning of science: A discourse focus. British Education Research Journal, 28(5), 673-688.  https://doi.org/10.1080/0141192022000015534
  4. Cao, Q., Lin, L., Shi, Y., Liang, X., & Li, G. (2017). Attention-aware face hallucination via deep reinforcement learning. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 690-698. 
  5. Chiang, T. T. C., Liao, C.-S., & Wang, W.-C. (2022a). Impact of artificial intelligence news source credibility identification system on effectiveness of media literacy education. Sustainability, 14, 4830. 
  6. Chiang, T. T. C., Liao, C.-S., & Wang, W.-C. (2022b). Investigating the difference of fake news source credibility recognition between ANN and BERT algorithms in artificial intelligence. Applied Sciences, 12, 7725. 
  7. Chowdhary, K. R. (2020). Fundamentals of artificial intelligence. Dordrecht: Springer. 
  8. Cooper, G. (2023). Examining science education in ChatGPT: An exploratory study of generative artificial intelligence. Journal of Science Education and Technology, 32, 444-452.  https://doi.org/10.1007/s10956-023-10039-y
  9. Copeland, J. (2015). Artificial intelligence: A philosophical introduction. Oxford: Blackwell. 
  10. Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018). BERT: Pre-training of deep bidirectional transformers for language understanding. Retrieved April 29, 2023 from https://arxiv.org/abs/1810.04805 
  11. Duit, R. (1991). On the role of analogies and metaphors in learning science. Science Education, 75(6), 649-672.  https://doi.org/10.1002/sce.3730750606
  12. Eisenstein, J. (2019). Introduction to natural language processing. Cambridge, MA: The MIT Press. 
  13. Fergus, S., Botha, M., & Ostovar, M. (2023). Evaluating academic answers generated using chatGPT. Journal of Chemical Education, 100, 1672-1675.  https://doi.org/10.1021/acs.jchemed.3c00087
  14. Fiori, A. (2019). Trends and applications of text summarization techniques. Hershey, PA: IGI Global. 
  15. Foster, D. (2023). Generative deep learning. Sebastopol, CA: O'Reilly.
  16. Goldenthal, E., Park, J., Liu, S. X., Mieczkowski, H., & Hancock, J. T. (2021). Not all AI are equal: Exploring the accessiblity of AI-mediated communication technology. Computers in Human Behavior, 125, 106975. 
  17. Goldie, J. G. S. (2016) Connectivism: A knowledge learning theory for the digital age? Medical Teacher, 38(10), 1064-1069.  https://doi.org/10.3109/0142159X.2016.1173661
  18. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning: Adaptive computation and machine learning series. Cambridge, MA: The MIT Press. 
  19. Han, S. (2023). Evolution of large language models and cloud services such as ChatGPT. Digital Service Issue Report, 3, 3-12. 
  20. Hinton, G. E., Osindero, S., & Teh, Y. (2006). A fast learning algorithm for deep belief nets. Neural Computation, 18, 1527-1554.  https://doi.org/10.1162/neco.2006.18.7.1527
  21. Hu, P., Lu, Y., Gong, Y. (2021). Dual humanness and trust in conversational AI: A person-centered approach. Computers in Human Behavior, 119, 106727. 
  22. Humphry, T., & Fuller, A. L. (2023). Potential ChatGPT use in undergraduate chemistry laboratories. Journal of chemical Education, 100, 1434-1436.  https://doi.org/10.1021/acs.jchemed.3c00006
  23. Jho, H. (2020). Discussion how to apply artificial intelligence to physics education. New Physics: Sae Mulli, 70(11), 974-984.  https://doi.org/10.3938/NPSM.70.974
  24. Jho, H., & Lee, B. (2022). Clustering science gifted students' graduation thesis based on machine learning. Journal of Science Education for the Gifted, 14(1), 13-22. 
  25. Kabudi, T., Pappas, I., & Olsen, D. H. (2021). AI-enabled adaptive learning system: A systematic mapping of the literature. Computers and Education: Artificial Intelligence, 2, 100017. 
  26. Kang, D. (2023). The advent of ChatGPT and the response of Korean language education. Korean Language and Literature, 82, 469-496. 
  27. Kim, S., Kim, S., Lee, M., & Kim, H. (2020). Review on artificial intelligence education for K-12 students and teachers. The Journal of Korean Association of Computer Education, 23(4), 1-11.  https://doi.org/10.32431/kace.2020.23.1.001
  28. Kizito, R. N. (2016). Connectivism in learning activity design: Implications for pedagogically-based technology adoption in African higher education contexts. International Review of Research in Open and Distributed Learning, 17(2), 19-39.  https://doi.org/10.19173/irrodl.v17i2.2217
  29. Ko, B., & Han, S. (2021). Achievements in AI education of elementary school teachers and awareness of AI education training. Korean Association of Artificial Intelligence Education Transaction, 2(1), 29-43. 
  30. Langr, J., & Bok, V. (2019). GANs in action: Deep learning with generative adversarial networks. New York: Manning. 
  31. Lee, J. (2023). Exploring the possibility of automatic scoring for graphical responses using a convolutional neural network. New Physics: Sae Mulli, 73(2), 138-149.  https://doi.org/10.3938/NPSM.73.138
  32. Lee, S., & Jeon, S. (2023). Issues about copyright of ChatGPT. Korean Copyright Commission. 
  33. Liu, B., Jiang, Y., Zhang, X., Liu, Q., Zhang, S., Biswas, J., & Stone, P. (2023). LLM+P: Empowering large language models with optimal planning proficiency. Retrieved May 1, 2023 from https://arxiv.org/abs/2304.11477 
  34. Liu, C., Shum, H.-Y., & Freeman, W. T. (2007). Face hallucination: Theory and practice. International Journal of Computer Vision, 75, 115-134.  https://doi.org/10.1007/s11263-006-0029-5
  35. Lu, Z., Pu, H., Wang, F., Hu, Z., & Wang, L. (2017). The expressive power of neural networks: A view from the width. Proceeding of the 31st conference on Neural Information Processing Systems, Long Beach, CA. 
  36. Many, I., & Maybury, M. T. (1999). Advances in automatic text summarization. Cambridge, MA: The MIT Press. 
  37. McCulloch, W. S., & Pitts, W. H. (1943). A log ical calculus of the ideas immanent in nervous activity. Bulletin of Biophysics, 5, 115-133.  https://doi.org/10.1007/BF02478259
  38. Park, G., Hwang, S., & Lee, J. (2023). Development and validation of teaching competence scale for teachers' artificial intelligence convergence education. Journal of Education Technology, 39(1), 315-344. 
  39. Partala, T., & Surakka, V. (2004). The effects of affective interventions in human-computer interaction. Interacting with Computers, 16(2), 295-309.  https://doi.org/10.1016/j.intcom.2003.12.001
  40. Peters, M. E., Neumann, M., Iyyer, M., Gardner, M., Clark, C., Lee, K., & Zettlemoyer, L. (2018). Deep contextualized word representations. Retrieved April 30, 2023 from https://arxiv.org/abs/1802.05365 
  41. Radford, A., Narasimhan, K., Salimans, T., & Sutskever, I. (2018). Improving language understanding by generative pre-training. Retrieved April 30, 2023 from https://s3-us-west-2.amazonaws.com/openai-assets/researchcovers/language-unsupervised/language_understanding_paper.pdf 
  42. Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2019). Exploring the limits of transfer learning with a unified text-to-text transformer. Retrieved April 25, 2023 from https://arxiv.org/abs/1910.10683 
  43. Raghu, M., Poole, B., Kleinberg, J., Ganguli, S., & Sohl-Dickstein, J. (2016). On the expressive power of deep neural networks. Retrived May 5, 2023 from https://arxiv.org/abs/1606.05336 
  44. Rapp, A., Curti, L., & Boldi, A. (2021). The human side of human-chatbot interaction: A systematic literature review of ten years of research on text-based chatbots. International Journal of Human-Computer Studies, 151, 102630. 
  45. Ravichandiran, S. (2019). Hands-on deep learning algorithms with python. Bermingham: Packt. 
  46. Rosenblatt, F. (1958). The Perceptron: A probabilistic model for information storage and organization in the brain. Psychological Review, 65(6), 386-408.  https://doi.org/10.1037/h0042519
  47. Shin, D., Jeong, H., & Lee, Y. (2023). Exploring the potential of using ChatGPT as a content-based English learning and teaching tool. Journal of the Korea English Education Society, 22(1), 171-192. 
  48. Shu, K., Wang, S., Lee, D., & Liu, H. (2020). Disinformation, misinformation, and fake news in social media. Dordrecht: Springer. 
  49. Shulman, J., Wolski, F., Dhariwal, P., Radford, A., & Klimov, O. (2017). Proximal policy optimization algorithms. Retrieved May 3, 2023 from https://arxiv.org/abs/1707.06347 
  50. Sutton, R. S., & Barto, A. G. (2018). Reinforcement learning: An introduction. Cambridge, MA: The MIT Press. 
  51. Sutskever, I., Vinyals, O., & Le, Q. V. (2014). Sequence to sequence learning with neural networks. Proceeding of the 27th International Conference on Neural Information Processing Systems, Montreal, Canada. 
  52. Tilli, A., Shehata, B., Adarkwah, M. A., Bozkurt, A., Hickey, D. T., Huang, R., & Agyemang, B. (2023). What if the devil is my guardian angel: ChatGPT as a case study of using chatbots in education. Smart Learning Environments, 10, 15. 
  53. Transue, B. M. (2013). Connectivism and information literacy: Moving from learning theory to pedagogical practice. Public Service Quarterly, 9(3), 185-195.  https://doi.org/10.1080/15228959.2013.815501
  54. Tunstall, L., von Werra, L., & Wolf, T. (2022). Natural language processing with transformers. Sebastopol, CA: O'Reilly. 
  55. UNESCO. (2023). ChatGPT and artificial intelligence in higher education: Quick start guide. Retrieved May 3, 2023 from https://unesdoc.unesco.org/ark:/48223/pf0000385146 
  56. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kalser, L., & Polosukhin, I. (2017). Attention is all you need. Retrieved April 3, 2023 from https://arxiv.org/abs/1706.03762 
  57. Wang, L., Hu, Y., He, J., Xu, X., Liu, N., Liu, H., & Shen, H. T. (2023). T-SciQ: Teaching multimodal chain-of-thought reasoning via large language model signals for science question answering. Retrieved May 31, 2023 from https://arxiv.org/abs/2305.03453 
  58. Weise, K., & Metz, C. (2023, May 9). When A.I. chatbots hallucinate. New York Times, https://www.nytimes.com/2023/05/01/business/aichatbots-hallucination.html 
  59. Xian, Y., Lampert, C. H., Schiele, B., & Akata, Z. (2020). Zero-shot learning: A comprehensive evaluation of the good, the bad and the ugly. Retrieved May 6, 2023 from https://arxiv.org/abs/1707.00600