THE RECOGNITION OF LONTARA SCRIPT IN THE MANDAR LANGUAGE USING CONVOLUTIONAL NEURAL NETWORK

Nurahdika, Ni Putu Irene Pasca (2025) THE RECOGNITION OF LONTARA SCRIPT IN THE MANDAR LANGUAGE USING CONVOLUTIONAL NEURAL NETWORK. Tugas Akhir thesis, Informatics.

Text 5210411081_Ni Putu Irene Pasca Nurahdika_Abstrak.pdf Download (67kB)

Abstract

The Mandar language, spoken by the Mandar ethnic group in West Sulawesi Province—particularly in the districts of Polewali Mandar, Majene, and Mamuju—uses the traditional Lontara script in everyday communication. Both the Mandar language and Lontara script are part of Indonesia’s rich cultural heritage that must be preserved through modern technological means. This study implements a Convolutional Neural Network (CNN) to recognize handwritten Lontara characters, consisting of 23 distinct symbols with various form variations. The Lontara dataset comprises 7,452 training images and 3,726 validation images. Preprocessing stages included normalization and data augmentation to enhance model generalization. The CNN architecture incorporates three convolutional layers, ReLU activation functions, dropout layers to mitigate overfitting, and a softmax output layer. The model is optimized using the Adam algorithm and categorical crossentropy loss function. Results show a training accuracy of 89.69% and validation accuracy of 80.76%, with a significant decrease in loss. Additionally, the study includes 500 bilingual vocabulary entries (Indonesian–Mandar) to support linguistic context. Graphical analysis reveals strong consistency between training and validation data, indicating good generalization capability. This research demonstrates the effectiveness of CNN in recognizing Lontara script and supports the digital preservation of cultural heritage. Keywords: Lontara Script, Mandar Language, Convolutional Neural Network, Pattern Recognition.

Item Type:	Thesis (Skripsi, Tugas Akhir or Kerja Praktek) (Tugas Akhir)
Subjects:	T Technology > T Technology (General)
Divisions:	Fakultas Sains Dan Teknologi > S1 Informatika
Depositing User:	Kaprodi S1 Informatika UTY
Date Deposited:	19 Jul 2025 01:19
Last Modified:	19 Jul 2025 01:19
URI:	http://eprints.uty.ac.id/id/eprint/18276

Actions (login required)

View Item