인공지능 기반 학습자 맞춤형 교육을 위한 형평성과 편향성 연구
Addressing Bias for Equity in AI-driven Customized Learning Systems
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The integration of AI into education has catalyzed transformative changes, particularly in enabling customized and personalized learning experiences that were previously impractical in traditional classroom settings. While earlier AI automation systems focused on educational equality by providing uniform resources and opportunities, more recent AIdriven customized learning systems aim to achieve educational equity by providing differentiated support tailored to individual needs, ensuring comparable learning outcomes across diverse student populations. However, these AI-driven customized learning systems can inadvertently introduce bias in both data collection and algorithmic processing, potentially compromising educational equity. This risk is particularly pronounced in foreign language education, where learner populations exhibit significant demographic and cultural diversity, increasing the potential for data bias. With the planned introduction of AI English digital textbooks in 2025, addressing AI bias and educational inequality becomes critical, particularly in English education. This study systematically examines the sources of AI bias that occur at multiple stages: data collection, analysis, classification, and algorithmic processing. It identifies the manifestations and implications of these biases in educational contexts, and proposes comprehensive solutions that include both technical and non-technical approaches, drawing on existing literature. It also outlines recommendations for key stakeholders -developers, educators, and policymakers- to ensure equity in AI-driven educational systems.
Abstract
The integration of AI into education has catalyzed transformative changes, particularly in enabling customized and personalized learning experiences that were previously impractical in traditional classroom settings. While earlier AI automation systems focused on educational equality by providing uniform resources and opportunities, more recent AIdriven customized learning systems aim to achieve educational equity by providing differentiated support tailored to individual needs, ensuring comparable learning outcomes across diverse student populations. However, these AI-driven customized learning systems can inadvertently introduce bias in both data collection and algorithmic processing, potentially compromising educational equity. This risk is particularly pronounced in foreign language education, where learner populations exhibit significant demographic and cultural diversity, increasing the potential for data bias. With the planned introduction of AI English digital textbooks in 2025, addressing AI bias and educational inequality becomes critical, particularly in English education. This study systematically examines the sources of AI bias that occur at multiple stages: data collection, analysis, classification, and algorithmic processing. It identifies the manifestations and implications of these biases in educational contexts, and proposes comprehensive solutions that include both technical and non-technical approaches, drawing on existing literature. It also outlines recommendations for key stakeholders -developers, educators, and policymakers- to ensure equity in AI-driven educational systems.
참고문헌 (52)
- Alesina, A., "Revealing stereotypes : Evidence from immigrants in schools", American Economic Review, 114(7), pp.1916~1948 (2024)[DOI]
- Baker, R. S., "Algorithmic bias in education", International Journal of Artificial Intelligence in Education, 32, pp.1052~1092 (2022)[DOI]
- Barbosa, G., "Towards automatic cross-language classification of cognitive presence in online discussions", Proceedings of the Tenth International Conference on Learning Analytics & Knowledge, 605-614. https :, doi. org, pp.10. 1145, 337546[DOI]
- Bridgeman, B., "Comparison of human and machine scoring of essays : Differences by gender, ethnicity, and country", Applied Measurement in Education, 25(1), pp.27~40 (2012)[DOI]
- Bulathwela, S., "Artificial intelligence alone will not democratise education : On educational inequality, techno-solutionism and inclusive tools", Sustainability, 16(2), pp.781~801 (2024)[DOI]
- Cavalcanti, A, "How good is my feedback? A content analysis of written feedback", Proceedings of the tenth international conference on learning analytics & knowledge, 428-437. https :, doi. org, pp.10. 1145, 337546[DOI]
- Chen, L, "Artificial intelligence in education: A review", IEEE Access, 8, pp.75264~75278 (2020)[DOI]
- Chouldechova, A., "A snapshot of the frontiers of fairness in machine learning", Communications of the ACM, 63(5), pp.82~89 (2020)[DOI]
- Darling-Hammond, L, "The right thing to do, the smart thing to do: Enhancing diversity in the health professions", pp.208~233 (2001), National Academy Press
- Deng, G., "Technological pedagogical content ethical knowledge (TPCEK): The development of an assessment instrument for pre-service teachers", Computers & Education, 197, pp.10474~ (2023)[DOI]
- Eden, C-A., "Integrating AI in education : Opportunities, challenges, and ethical considerations", Magna Scientia Advanced Research and Review, 10(2), pp.6~13 (2024)[DOI]
- "The unintended consequences of artificial intelligence and education" (2023)[URL]
- Holstein, K., "Co-designing a real-time classroom orchestration tool to support teacher-AI complementarity", Journal of Learning Analytics, 6(2), pp.27~52 (2019)[DOI]
- Holstein, K., "Improving fairness in machine learning systems: What do industry practitioners need?", Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems, 600 (2019)
- Hong, S., "Expectations and concerns about adopting AI digital textbooks : Based on investigation of teachers’use of AI and digital tools", The Journal of Studies in Language, 40(1), pp.7~20 (2024)[DOI]
- Ibrahim, A., "The challenges of leveraging threat intelligence to stop data breaches", Frontier in Computer Science, 28, pp.1~11 (2020)[DOI]
- Idowu, J., "Investigating algorithmic bias in student progress monitoring", Computers and Education : Artificial Intelligence, 7, pp.100267~ (2024)[DOI]
- "Towards equity and algorithmic fairness in student grade prediction" (2021)[URL]
- Kasneci, E., "ChatGPT for good? On opportunities and challenges of large language models for education", Learning and Individual Differences, 103, pp.102274~ (2023)[DOI]
- Kim, S., "One more way of understanding the education in the era of AI", The Journal of Educational Principles, 24(1), pp.83~105 (2019)[DOI]
- Kim, S., "The introduction of the concept of ‘Education AI’ : The challenge of sustainable education by the cooperation between human and AI", The Journal of Educational Principles, 25(1), pp.1~21 (2020)[DOI]
- Kim, S., "A discussion on the planning of national digital transformation in the education sector", The Journal of Korea Education, 45(4), pp.173~200 (2018)[DOI]
- Knox, J., "Machine behaviourism: Future visions of ‘learnification’ and ‘datafication’ across humans and digital technologies", Learning, Media and Technology, 45(1), pp.31~45 (2019)[DOI]
- Kwak, Y., "Bridging large language model disparities : Skill tagging of multilingual educational content", British Journal of Educational Technology, 55(5), pp.2039~2057 (2024)[DOI]
- Ladson-Billings, G., "From the achievement gap to the education debt: Understanding achievement in U.S. schools", Educational Researcher, 35(7), pp.3~12 (2006)
- Lee, M. J., "The landscape and gaps in open source fairness toolkits", Proceedings of the 2021CHI Conference on Human Factors in Computing Systems, 699 (2021)
- Meaney, M, "Early adopter iteration bias and research praxia bias in the learning analytic ecosystem", Learning Analytics and Knowledge 2019: Companion Proceedings of the 9th International Learning Analytics & Knowledge Conference, pp.14~20 (2019)
- "Realizing personalized education for all: Strategies for digital-based educational innovation" (2023)[URL]
- "Accurate measurement of lexical sophistication with reference to ESL learner data" (2018)[URL]
- Park, Y., "A grounded theory on AI-based teacher supporting platform", Journal of Korean Association for Educational Information and Media, 30(3), pp.1005~1034 (2024)[DOI]
- "AI can disrupt racial inequity in schools, or make it much worse" (2019)[URL]
- Plaza P., "Educational robotics for all:Gender, diversity, and inclusion in steam", Proceedings of IEEE Learn, pp.19~24 (2020), MOOCS
- Ramineni, C., "Understanding mean score differences between the e-rater R automated scoring engine and humans for demographically based groups in the GRER general test", ETS Research Report Series, 2, pp.1~31 (2018)[DOI]
- "From good intentions to real outcomes: Equity by design in learning technologies" (2017)[URL]
- Sha, L., "Leveraging class balancing techniques to alleviate algorithmic bias for predictive tasks in education", IEEE Transactions on learning and Technologies, 15(4), pp.481~492 (2022)[DOI]
- Sha, L., "Lecture Notes in Computer Science", pp.12748~ (2021), Springer
- Smith, M., "Students’ views on fairness in education : The importance of relational justice and stakes fairness", Research Papers in Education, 33(3), pp.336~353 (2017)[DOI]
- Stray, J., "The AI learns to lie to please you : Preventing biased feedback loops in machine-assisted intelligence analysis", Analytics, 2(2), pp.350~358 (2023)[DOI]
- Suresh, H., "A framework for understanding sources of harm throughout the machine learning life cycle", Proceedings of the 1st ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization, 17, pp.1~9 (2021)
- Unterhalter, E., "What is equity in education? Reflections from the capability approach", Stud Philos Educ, 28, pp.415~424 (2009)[DOI]
- Varsik, S., "The potential impact of artificial intelligence on equity and inclusion in education: OECD artificial intelligence papers", pp.1787~ (2024), OECD Publishing
- Wang, Z., "Monitoring the performance of human and automated scores for spoken responses", Language Testing, 35(1), pp.101~120 (2018)[DOI]
- Warschauer, M., "The affordances and contradictions of AI-generated text for writers of English as a second or foreign language", Journal of Second Language Writing, 62, pp.101071~ (2023)[DOI]
- Wong, D, "Examining the relationship between student attributional style, perceived teacher fairness, and sense of school belonging", Contemporary Educational Psychology, 71, pp.102113~ (2022)[DOI]
- Yang, Y, "Enhancing fairness in face detection in computer vision systems by demographic bias mitigation", Proceedings of the 2022 AAAI, 2022, 813~822
- Yu, H., "Towards AI-powered personalization in MOOC learning", npj Science of Learning, 2, pp.15~ (2017)[DOI]
- "Enhancing fairness and performance in machine learning models: A multi-task learning approach with Monte-Carlo dropout and Pareto optimality" (2024)[URL]
- Zou, J., "AI can be sexist and racist — it’s time to make it fair", Nature, 559(7714), pp.324~326 (2018)[DOI]