AUTOMATED ESSAY SCORING UNTUK PENILAIAN JAWABAN ESAI BAHASA INDONESIA DENGAN INDOBERT EMBEDDING DAN FEEDFORWARD NEURAL NETWORK

HUMAIRA, RAMADHANIA and Abdiansah, Abdiansah (2025) AUTOMATED ESSAY SCORING UNTUK PENILAIAN JAWABAN ESAI BAHASA INDONESIA DENGAN INDOBERT EMBEDDING DAN FEEDFORWARD NEURAL NETWORK. Undergraduate thesis, Sriwijaya University.

	Image RAMA_55201_09021282126064_cover.jpg - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (481kB)
	Text RAMA_55201_09021282126064.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (3MB) | Request a copy
	Text RAMA_55201_09021282126064_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7MB) | Request a copy
	Text RAMA_55201_09021282126064_0001108401_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (1MB)
	Text RAMA_55201_09021282126064_0001108401_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (487kB) | Request a copy
	Text RAMA_55201_09021282126064_0001108401_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (359kB) | Request a copy
	Text RAMA_55201_09021282126064_0001108401_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (863kB) | Request a copy
	Text RAMA_55201_09021282126064_0001108401_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_55201_09021282126064_0001108401_06.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (183kB) | Request a copy
	Text RAMA_55201_09021282126064_0001108401_07_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (194kB) | Request a copy
	Text RAMA_55201_09021282126064_0001108401_08_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (131kB) | Request a copy

Abstract

Improving the quality of education can be supported by a more effective assessment system, one of which is Automated Essay Scoring (AES) for automatic essay evaluation. This study develops an Indonesian-language AES system using IndoBERT Embedding and a Feedforward Neural Network (FNN). The dataset used is the secondary dataset from the UKARA Challenge, developed by the NLP Research Team at Universitas Gadjah Mada, which has a limited number of data and an imbalanced class distribution (labels 1 and 0). Overall, the developed model, after being trained and evaluated on datasets A and B, achieved an F1-score of 0.767. On dataset A, the model trained using the SMOTE technique obtained an F1-score of 0.835 with a batch size of 16, epoch 7, and a learning rate of 1.26e-4. The best model on dataset B achieved an F1-score of 0.699 with a batch size of 64, epoch 4, and a learning rate of 5.7e-3. These results indicate that IndoBERT Embedding and FNN provide a reasonably good performance compared to the baseline provided by UKARA for the training set, although challenges remain regarding data imbalance and limited dataset size.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	AES, IndoBERT, FNN, SMOTE
Subjects:	T Technology > T Technology (General) > T1-995 Technology (General)
Divisions:	09-Faculty of Computer Science > 55201-Informatics (S1)
Depositing User:	Ramadhania Humaira
Date Deposited:	14 Mar 2025 01:53
Last Modified:	14 Mar 2025 01:53
URI:	http://repository.unsri.ac.id/id/eprint/168660

Actions (login required)

View Item