@article{ART003093035},
author={Maresha Caroline Wijanto and Hwanseung Yong},
title={Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data},
journal={Journal of Software Assessment and Valuation},
issn={2092-8114},
year={2024},
volume={20},
number={2},
pages={35-45}
TY - JOUR
AU - Maresha Caroline Wijanto
AU - Hwanseung Yong
TI - Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data
JO - Journal of Software Assessment and Valuation
PY - 2024
VL - 20
IS - 2
PB - Korea Software Assessment and Valuation Society
SP - 35
EP - 45
SN - 2092-8114
AB - The automatic grading of short answer question is important in the field of Natural Language Processing. ASAG (Automated Short Answer Grading) task have undergone numerous advancements.
Recent studies have adopted transformer models such as the T5 embedding or BERT-base models.
Nonetheless, ASAG tasks encounter significant challenges stemming from limited data availability. The urgent need for more training data emerges as a central issue. Several researchers have proposed augmentation approaches to address this gap. In this study, we introduce other data augmentation technique utilizing prompt engineering by the GPT model. We deploy ASAG system using the Sentence Transformers model, fine-tuning specific hyper-parameters alongside the augmented dataset. The primary factors influencing performance enhancement include the augmentation process, particularly the quantity of augmented data, and the dataset split size for training and testing purposes. Furthermore, alternative GPT models or fine-tuning GPT could be explored within the augmentation process.
KW - Data Augmentation;Automated Short Answer Grading System;GPT;fine-tuning
DO -
UR -
ER -
Maresha Caroline Wijanto and Hwanseung Yong. (2024). Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data. Journal of Software Assessment and Valuation, 20(2), 35-45.
Maresha Caroline Wijanto and Hwanseung Yong. 2024, "Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data", Journal of Software Assessment and Valuation, vol.20, no.2 pp.35-45.
Maresha Caroline Wijanto, Hwanseung Yong "Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data" Journal of Software Assessment and Valuation 20.2 pp.35-45 (2024) : 35.
Maresha Caroline Wijanto, Hwanseung Yong. Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data. 2024; 20(2), 35-45.
Maresha Caroline Wijanto and Hwanseung Yong. "Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data" Journal of Software Assessment and Valuation 20, no.2 (2024) : 35-45.
Maresha Caroline Wijanto; Hwanseung Yong. Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data. Journal of Software Assessment and Valuation, 20(2), 35-45.
Maresha Caroline Wijanto; Hwanseung Yong. Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data. Journal of Software Assessment and Valuation. 2024; 20(2) 35-45.
Maresha Caroline Wijanto, Hwanseung Yong. Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data. 2024; 20(2), 35-45.
Maresha Caroline Wijanto and Hwanseung Yong. "Modeling Short Answer Grading Performance Improvement by GPT Augmentation Data" Journal of Software Assessment and Valuation 20, no.2 (2024) : 35-45.
