Breadcrumb
- Home
- AI Hub For Education
- Research Study Repository
- Evidence Explorer
- AI Tools
- Custom-built Research Tool
The Repository
Search and Filter
Custom-built research tool
Studies tagged with AI from their full text or abstract; see each study for details.
Showing 421 - 450 of 531 studies
Young Children's Anthropomorphism Of An Ai Chatbot: Brain Activation And The Role Of Parent Co-Presence
Pilyoung Kim. (03/2025). arXiv.
Children attributed humanlike qualities across all domains to both partners, rating parents higher overall but giving the AI relatively high perceptive and epistemic scores.
VTutor: An Open-Source SDK for Generative AI-Powered Animated Pedagogical Agents with Multi-Media Output
Eason Chen, Chenyu Lin, Xinyi Tang, Aprille Xi, Canwen Wang, Jionghao Lin, Kenneth R. Koedinger. (02/2025). arXiv.
VTutor provides an open-source development kit that links large language models, text-to-speech audio, and real-time lip synchronization to generate dynamic animated agents.
The Art of Storytelling: Multi-Agent Generative AI for Dynamic Multimodal Narratives
Samee Arif, Taimoor Arif, Muhammad Saad Haroon, Aamina Jamal Khan, Agha Ali Raza, Awais Athar. (02/2025). arXiv.
Gemma-2-9b ranked highest overall among evaluated language models for story generation quality, outperforming larger models like Gemma-2-27b and GPT-4o.
SocratiQ: A Generative AI-Powered Learning Companion for Personalized Education and Broader Accessibility
Jason Jabbour, Kai Kleinbard, Olivia Miller, Robert Haussman, Vijay Janapa Reddi. (02/2025). arXiv.
Analysis of 100 beginner-level questions revealed lower-order cognitive skills dominated, with 42% targeting remembering and 28% understanding, while applying, analyzing, evaluating, and creating comprised 14%, 7%, 3%, and 2% respectively.
SET-PAIRED: Designing for Parental Involvement in Learning with an AI-Assisted Educational Robot
Hui-Ru Ho, Nitigya Kargeti, Ziqi Liu, Bilge Mutlu. (02/2025). arXiv.
Parents dynamically varied their collaboration modes between full parent takeover, shared involvement, and robot takeover depending on their available time, energy, and trust in AI.
MindCraft: Revolutionizing Education through AI-Powered Personalized Learning and Mentorship for Rural India
Arihant Bardia, Aayush Agrawal. (02/2025). arXiv.
In a rural Madhya Pradesh case study, six months of platform use was associated with a 40-point increase in the student's mathematics and English exam scores.
One Size doesn't Fit All: A Personalized Conversational Tutoring Agent for Mathematics Instruction
Ben Liu, Jihai Zhang, Fangquan Lin, Xu Jia, Min Peng. (02/2025). arXiv.
PACE outperformed baseline language models on reference-based evaluation metrics, achieving an average score of 34.55 compared to 19.97 to 27.70 for baselines.
EdgeAIGuard: Agentic LLMs for Minor Protection in Digital Spaces
Ghulam Mujtaba, Sunder Ali Khowaja, Kapal Dev. (02/2025). arXiv.
EdgeAIGuard achieved 93.4% threat detection accuracy and a 92.9% F1 score, outperforming existing baseline models.
TutorLLM: Customizing Learning Recommendations with Knowledge Tracing and Retrieval-Augmented Generation
Zhaoxing LI, Vahid Yazdanpanah, Jindi Wang, Wen Gu, Lei Shi, Alexandra Ioana Cristea, Sarah Kiden, Sebastian Stein. (01/2025). arXiv.
Differences in student exam performance across groups were not statistically significant, though average scores were highest for TutorLLM (74.48) compared to hybrid (72.81) and general LLM groups (71.97).
A Novel Approach to Scalable and Automatic Topic-Controlled Question Generation in Education
Ziqing Li, Mutlu Cukurova, Sahan Bulathwela. (01/2025). arXiv.
Fine-tuning a small T5 model on contrastive data (TopicQG) generated more linguistically accurate and topic-aligned questions than the baseline model.
Auto-Evaluation: A Critical Measure in Driving Improvements in Quality and Safety of AI-Generated Lesson Resources
Hannah-Beth Clark, Margaux Dowland, Laura Benton, Reka Budai, Ibrahim Kaan Keskin, Emma Searle, Matthew Gregory, Mark Hodierne, William Gayne, John Roberts. (01/2025). arXiv.
Incorporating teacher guidance and few-shot examples into the evaluation prompt significantly reduced the mean squared error between automated and human scores from 3.83 to 2.95.
Enhancing textual textbook question answering with large language models and retrieval augmented generation
Hessa Abdulrahman Alawwad, Areej Alhothali, Usman Naseem, Ali Alkhathlan, Amani Jamal. (01/2025). arXiv.
The proposed framework achieved 84.24% overall accuracy on the test set, outperforming prior state-of-the-art textbook question answering models.
Exploring Iterative Enhancement for Improving Learnersourced Multiple-Choice Question Explanations with Large Language Models
Qiming Bao, Juho Leinonen, Alex Yuxuan Peng, Wanjun Zhong, Gaël Gendron, Tim Pistotti, Alice Huang, Paul Denny, Michael Witbrock, Jiamou Liu. (01/2025). arXiv.
The iterative ILearner-LLM framework produced explanations closer to student-written text than single-generation baselines, outperforming fine-tuned LLaMA2-13B by 5.33 and GPT-4 by 3.86 in BLEU scores.
Generative AI in Education: From Foundational Insights to the Socratic Playground for Learning
Xiangen Hu, Sheng Xu, Richard Tong, Art Graesser. (01/2025). arXiv.
The Socratic Playground for Learning architecture provides five adaptive modes—Assessment, Tutoring, Vicarious, Gaming, and Teachable Agent—to support progressive, personalized student inquiry.
iLLuMinaTE: An LLM-XAI Framework Leveraging Social Science Explanation Theories Towards Actionable Student Performance Feedback
Vinitra Swamy, Davide Romano, Bhargav Srinivasa Desikan, Oana-Maria Camburu, Tanja Kaser. (01/2025). arXiv.
University students preferred the theory-driven iLLuMinaTE explanations over traditional baseline explainers 89.52% of the time.
IntelliChain: An Integrated Framework for Enhanced Socratic Method Dialogue with LLMs and Knowledge Graphs
Changyong Qi, Linzhao Jia, Yuang Wei, Yuan-Hao Jiang, Xiaoqing Gu. (01/2025). arXiv.
Integrating teaching agents with a knowledge graph improved dialogue quality, deepening conceptual discussion and guiding algebraic problem-solving through the Socratic method.
Harnessing AI in Secondary Education to Enhance Writing Competence
Eyvind Elstad, Harald Eriksen. (12/2024). arXiv.
Short and highly specific prompts allow AI to provide concrete, targeted feedback that helps students revise text structure without losing ownership of their writing.
AI-Driven Virtual Teacher for Enhanced Educational Efficiency: Leveraging Large Pretrained Models for Autonomous Error Analysis and Correction
Tianlong Xu, Yi-Fan Zhang, Zhendong Chu, Shen Wang, Qingsong Wen. (12/2024). arXiv.
Expert evaluation showed that the VATE system achieved 78.3% accuracy in analyzing student error causes and generating learning recommendations on challenging problems.
Al-Supported Data Analysis Boosts Student Motivation and Reduces Stress in Physics Education
Jannik Henze, André Bresges, Sebastian Becker-Genschow. (12/2024). arXiv.
Both groups showed significant pre-to-post learning gains, but there were no statistically significant differences in cognitive performance between the AI and Excel groups.
ActiveAI: Enabling K-12 AI Literacy Education & Analytics at Scale
Ruiwei Xiao, Ying-Jui Tseng, Hanqi Li, Hsuan Nieu, Guanze Liao, John Stamper, Kenneth R. Koedinger. (12/2024). arXiv.
Students achieved significant learning gains in four of seven learning objectives across four instructional modules.
Beware of Metacognitive Laziness: Effects of Generative Artificial Intelligence on Learning Motivation, Processes, and Performance
Yizhou Fan, Luzhen Tang, Huixiao Le, Kejie Shen, Shufang Tan, Yueying Zhao, Yuan Shen, Xinyu Li, Dragan Gašević. (12/2024). arXiv.
Students supported by ChatGPT achieved significantly greater essay score improvements than those supported by a human expert, writing analytics tools, or working alone.
Tuning Music Education: AI-Powered Personalization in Learning Music
Mayank Sanganeria, Rohan Gala. (12/2024). arXiv.
Automatic Chord Recognition and beat detection enable an ear training app to generate personalized quizzes directly from user-selected audio tracks.
Revolutionizing Undergraduate Learning: CourseGPT and Its Generative AI Advancements
Ahmad M. Nazar, Mohamed Y. Selim, Ashraf Gaffar, Shakil Ahmed. (12/2024). arXiv.
The larger Mixtral-8x7b model outperformed the smaller Mistral-7b model, achieving an 88.0% correctness score compared to 78.2%.
Promoting AI Literacy in Higher Education: Evaluating the IEC-V1 Chatbot for Personalized Learning and Educational Equity
Stefan Pietrusky. (12/2024). arXiv.
Prospective teachers rated the ability to customize response proficiency levels highest, giving it a mean score of 4.14 out of 5.
MindScratch: A Visual Programming Support Tool for Classroom Learning Based on Multimodal Generative AI
Yunnong Chen, Shuhong Xiao, Yaxuan Song, Zejian Li, Lingyun Sun, Liuqing Chen. (12/2024). arXiv.
All 24 students completed teacher-defined objectives using MindScratch, compared to only 13 students when using Scratch.
LLMs as Debate Partners: Utilizing Genetic Algorithms and Adversarial Search for Adaptive Arguments
Prakash Aryan. (12/2024). arXiv.
Across 23 debates, the DebateBrawl system achieved an average score of 2.72 out of 10 compared to 2.67 for human debaters.
LearnLM: Improving Gemini for Learning
LearnLM Team, Google. (12/2024). arXiv.
Expert raters substantially preferred the LearnLM model across diverse learning scenarios compared to several leading AI models.
Breaking the Programming Language Barrier: Multilingual Prompting to Empower Non-Native English Learners
James Prather, Brent N. Reeves, Paul Denny, Juho Leinonen, Stephen MacNeil, Andrew Luxton-Reilly, João Orvalho, Amin Alipour, Ali Alfageeh, Thezyrie Amarouche, Bailey Kimmel, Jared Wright, Musa Blake, Gweneth Barbre. (12/2024). arXiv.
Most students successfully solved coding tasks prompting primarily in their native language, though native-language prompt success was higher for Portuguese (18%) and Chinese (15%) than Arabic (5%).
"From Unseen Needs to Classroom Solutions": Exploring AI Literacy Challenges & Opportunities with Project-Based Learning Toolkit in K-12 Education
Hanqi Li, Ruiwei Xiao, Hsuan Nieu, Ying-Jui Tseng, Guanze Liao. (12/2024). arXiv.
Over 92% of interviewed teachers believe students should use AI to support learning, though only 38% reported strong familiarity with the AI4K12 framework.
VISTA: Visual Integrated System for Tailored Automation in Math Problem Generation Using LLM
Jeongwoo Lee, Kwangsuk Park, Jihyeon Park. (11/2024). arXiv.
The multi-agent framework significantly outperformed the baseline model across all geometry and function problem subtypes in text coherence and problem relevance.
