Search Results for author:
Found 11 papers, 0 papers with code
How Can I Improve? Using GPT to Highlight the Desired and Undesired Parts of Open-ended Responses
To quantify the quality of highlighted praise components identified by GPT models, we introduced a Modified Intersection over Union (M-IoU) score.
Predicting Learning Performance with Large Language Models: A Study in Adult Literacy
This research is motivated by the potential of LLMs to predict learning performance based on its inherent reasoning and computational capabilities.
Improving Assessment of Tutoring Practices using Retrieval-Augmented Generation
The results indicate that the RAG prompt demonstrated more accurate performance (assessed by the level of hallucination and correctness in the generated assessment texts) and lower financial costs than the other strategies evaluated.
3DG: A Framework for Using Generative AI for Handling Sparse Learner Performance Data From Intelligent Tutoring Systems
Learning performance data (e. g., quiz scores and attempts) is significant for understanding learner engagement and knowledge mastery level.
Using Large Language Models to Assess Tutors' Performance in Reacting to Students Making Math Errors
By analyzing 50 real-life tutoring dialogues, we find both GPT-3. 5-Turbo and GPT-4 demonstrate proficiency in assessing the criteria related to reacting to students making errors.
Elucidating STEM Concepts through Generative AI: A Multi-modal Exploration of Analogical Reasoning
Our study demonstrates the potential of applying large language models to educational practice on STEM subjects.
AI Chatbots as Multi-Role Pedagogical Agents: Transforming Engagement in CS Education
This study investigates the use of Artificial Intelligence (AI)-powered, multi-role chatbots as a means to enhance learning experiences and foster engagement in computer science education.
Comparative Analysis of GPT-4 and Human Graders in Evaluating Praise Given to Students in Synthetic Dialogues
We found that both zero-shot and few-shot chain of thought approaches yield comparable results.
Using Large Language Models to Provide Explanatory Feedback to Human Tutors
Research demonstrates learners engaging in the process of producing explanations to support their reasoning, can have a positive impact on learning.
Robust Educational Dialogue Act Classifiers with Low-Resource and Imbalanced Datasets
We note that many prior studies on classifying educational DAs employ cross entropy (CE) loss to optimize DA classifiers on low-resource data with imbalanced DA distribution.
Does Informativeness Matter? Active Learning for Educational Dialogue Act Classification
Then, the study investigates how the AL methods can select informative samples to support DA classifiers in the AL sampling process.
