book icon

Reading List: Machine Learning and Artificial Intelligence in Education – A Critical Perspective

by

Edited by: Pati Ruiz, Aditi Mallavarapu, and Arun Balajiee

The following reading list was compiled to provide an overview of the current state of Machine Learning (ML) and Artificial Intelligence (AI) in education, also referred to as AI in Education Research (AIED). Some articles are focused broadly on research themes in AI/ML as they are applied for teaching and learning scenarios, while others more specifically on the intersection of equity, bias, ethics, of using ML/AI methods and the mechanics of the methods themselves. An additional reading section identifies relevant news articles, reports, and a podcast on these topics.

Included under each article is a brief bulleted overview of what each article covers and how much technical background each article requires.

Last Updated: 9/19/2022

Overview

Algorithmic Bias in Education
Citation: Baker, R. S., & Hawn, A. (2021). Algorithmic Bias in Education. International Journal of Artificial Intelligence in Education. https://doi.org/10.1007/s40593-021-00285-9
Background Needed: Low/Medium
Covers:

  • This review focuses on solidifying the current understanding of the concrete impacts of algorithmic bias in education.
  • The authors discuss theoretical and formal perspectives on algorithmic bias and review the evidence around algorithmic bias in education.
  • Finally, the authors propose a framework for moving from unknown bias to known bias and from fairness to equity and discuss obstacles to addressing these challenges and propose four areas for mitigating and resolving the problems of algorithmic bias in AIED systems and other educational technology.

Academic Articles

Culture in Computer-Based Learning Systems: Challenges and Opportunities
Citation: Baker, R. S., Walker, E., Ogan, A., & Madaio, M. (2020). Culture in Computer-Based Learning Systems: Challenges and Opportunities. https://doi.org/10.5281/zenodo.4057223
Background Needed: Medium
Covers:

  • The paper reviews various learning theories related to culture that can be used to predict learning behavior.
  • Identifies that it is not necessarily true that learners from collectivistic cultures would participate in group collaborative behavior since some results also indicate that learners from individualistic cultures also participate well in group discussions.
  • Compares and contrasts Hofstede’s model of culture with a simpler Inglehart-Welzel model as a trade-off between simplicity and degree of validation of individual values/individualism. Also, discusses Culture Based Learning (while there is no set definition, the paper cites literature that discusses Culture Based Learning).
  • The paper authors also acknowledge their own cultural background in the context of this discussion and that their perspectives may or may not be biased because of this background.

Algorithmic Fairness in Education
Citation: Kizilcec, R. F. & Lee, H. (2022). Algorithmic Fairness in Education. In W. Holmes
& K. Porayska-Pomsta (Eds.), Ethics in Artificial Intelligence in Education, Taylor & Francis. https://arxiv.org/abs/2007.05443
Background Needed: Medium
Covers:

  • Draws parallels to prior literature on educational access, bias, and discrimination.
  • Examines core components of algorithmic systems (measurement, model learning, and action) to identify sources of bias and discrimination in the process of developing and deploying these systems.
  • Provides recommendations for policy makers and developers of educational technology offer guidance for how to promote algorithmic fairness in education.

Unwritten Magic: Participatory Design of AI Dialogue to Empower Marginalized Voices
Citation: Buddemeyer A., Nwogu J., Solyst J., Walker E., Nkrumah T., Ogan A., Hatley L., and Stewart A (2022). Unwritten Magic: Participatory Design of AI Dialogue to Empower Marginalized Voices. https://doi.org/10.1145/3524458.3547119
Background Needed: Medium
Covers:

  • Participatory Design of dialogue systems with students of color in middle school.
  • One-week Youth Advisory Group (YAG) at a middle school in Washington D.C., to learn programming in Scratch.
  • Understand the linguistic features needed to develop a cultural-sensitive dialogue system.

Towards a Tripartite Research Agenda: A Scoping Review of Artificial Intelligence in Education Research
Citation: Wan, T., & Cheng, E. C. K. (2022). Towards a Tripartite Research Agenda: A Scoping Review of Artificial Intelligence in Education Research. In E. C. K. Cheng, R. B. Koul, T. Wang, & X. Yu (Eds.), Artificial Intelligence in Education: Emerging Technologies, Models and Applications (pp. 3–24). Springer Singapore.
Background Needed: Low/Medium
Covers:

  • This paper reviews research studies on artificial intelligence in education (AIED) published from 2001 to 2021
  • 135 manuscripts meeting the selection criteria were analyzed and three primary research areas were identified for AI in education research (AIED):
    • Learning from AI
    • Learning about AI, and
    • Learning with AI

On the genealogy of machine learning datasets: A critical history of ImageNet
Citation: Denton, E., Hanna, A., Amironesei, R., Smart, A., & Nicole, H. (2021). On the genealogy of machine learning datasets: A critical history of ImageNet. Big Data & Society. https://doi.org/10.1177/20539517211035955
Background Needed: Medium
Covers:

  • This paper conceptualizes machine learning datasets as a type of informational infrastructure in response to growing concerns of bias, discrimination, and unfairness perpetuated by algorithmic systems.
  • The paper examines the norms, values, and assumptions embedded in machine learning datasets and examines the histories and modes of constitution at play in their creation.
  • The researchers trace the discourses around large computer vision datasets and contribute to the development of the standards and norms around data development in machine learning and artificial intelligence research.

Evolution and Revolution in Artificial Intelligence in Education
Citation: Roll, I., & Wylie, R. (2016). Evolution and Revolution in Artificial Intelligence in Education. International Journal of Artificial Intelligence in Education, 26(2), 582–599. Evolution and Revolution in Artificial Intelligence in Education | SpringerLink
Background Needed: None/Low
Covers:

  • The authors take a historical perspective to identify the past foci that occupy the field of AIED. They consider 47 papers from three crucial years (early, middle, and recent years of AIED) in the history of the Journal of AIED (1994, 2004, and 2014).
  • The historical perspective considers focus of research (e.g., modeling, system review, evaluation), domain (e.g., STEM, language learning), types of problems (e.g. sequential problems, complex problems), collaboration structures (number of learners and machines), technology (e.g., computers, hand-helds), settings (e.g. school, workplace) and learning goals of each of the works.
  • They map this historical trajectory to current developments along the three dimensions of (1) Goals, (2) Practices and (3) Environments in AI research for education.

A Review of Artificial Intelligence (AI) in Education from 2010 to 2020
Citation: Zhai, X., Chu, X., Chai, C. S., Jong, M. S. Y., Istenic, A., Spector, M., Liu, J. B., Yuan, J., & Li, Y. (2021). A Review of Artificial Intelligence (AI) in Education from 2010 to 2020. Complexity, 2021, 539–547. https://doi.org/10.1155/2021/8812542
Background Needed: Medium (technical jargon)
Covers:

  • This paper presents a content analysis of 100 papers where artificial intelligence (AI) has been applied to the education sector.
  • Explores research trends as well as challenges for education that may be caused by AI including inappropriate use of AI techniques, changing roles of teachers and students, as well as social and ethical issues.
  • The authors define three layers of research interests: development layer – which includes mostly the methodology (classification, matching, recommendation, and deep learning), application layer which includes the end goal of applying the methods (feedback, reasoning, and adaptive learning), and integration layer that connects the computational elements in the methods back to the learning processes and targets (affection computing, role-playing, immersive learning, and gamification).
  • The authors also discuss challenges caused by AI due to inappropriate use of AI techniques, changing roles of teachers and students, and consideration of social and ethical issues along these perspectives.

Additional Readings and Resources

News and Magazine Articles

Datasheets for Datasets
Citation: Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., III, H. D., & Crawford, K. (2021). Datasheets for Datasets. Communications of the ACM, 64(12), 86–92. https://doi.org/10.1145/3458723
Background Needed: None
Covers:

  • Identifies issues with machine learning models that can reduce or amplify unwanted societal biases embedded in their training datasets.
  • Describes datasheets as a way to address the gap in documenting machine learning datasets by documenting the contexts and contents of datasets including: motivation; composition; collection processes; and recommended users.

The problems AI has today go back centuries
Citation: Hao, K. (2020, July 21). The problems AI has today go back centuries. MIT Technology Review. https://www.technologyreview.com/2020/07/31/1005824/decolonial-ai-for-everyone/
Background Needed: None

How to solve AI’s inequality problem
Citation: Rotman, D. (2022, April 19). How to solve AI’s inequality problem. MIT Technology Review. https://www.technologyreview.com/2022/04/19/1049378/ai-inequality-problem/
Background Needed: None

Reports

AI and education: Guidance for policy-makers
Citation: Miao, F., Holmes, W., Ronghuai Huang, & Hui Zhang. (n.d.). AI and education: Guidance for policy-makers. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000376709
Background Needed: Low

Trustworthy artificial intelligence (AI) in education (No. 218)
Citation: Vincent-Lancrin, S., & Vlies, R. van der. (2020). Trustworthy artificial intelligence (AI) in education (No. 218). OECD Publishing. https://doi.org/10.1787/a6c90fa9-en
Background Needed: Low

Podcasts

AI in Education Podcast
Citation: Bowen, D., Worrall, B., & Hickin, L. (n.d.). AI in Education Podcast. https://aipodcast.education/
Background Needed: None

Books

You Look Like a Thing and I Love You: How Artificial Intelligence Works and Why It’s Making the World a Weirder Place
Citation: Shane, J. (2019). You Look Like a Thing and I Love You: How Artificial Intelligence Works and Why It’s Making the World a Weirder Place. Little, Brown.
Background Needed: None