Skripsi

PENGARUH TEKNIK OVERSAMPLING DALAM KLASIFIKASI BAHASA DAERAH MENGGUNAKAN METODE LONG SHORT-TERM MEMORY (LSTM)

---

Regional language classification is one of the challenges in natural language processing due to the limited amount of data and the high lexical similarity among languages. This condition leads to data imbalance, particularly in minority classes, which can negatively affect model performance. This research aims to analyze the impact of oversampling techniques on the performance of the Long Short-Term Memory (LSTM) model for regional language classification. The oversampling techniques employed include text-based oversampling and feature-based oversampling using the Synthetic Minority Oversampling Technique (SMOTE). Word representations are constructed using FastText to transform words into numerical vectors before being processed by the LSTM model. The best model is obtained in the SMOTE scenario with a configuration of 256 LSTM units, a dropout rate of 0.3, a learning rate of 0.001, 10 epochs, and a batch size of 64. This model achieves an accuracy of 0.9714, with precision, recall, and f1-score values of 0.9590, 0.9623, and 0.9605, respectively.

---

Availability

Detail Information

Series Title

-

Call Number

T1929762026

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2026

Collation

xv, 202 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

006.350 7

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Teknik Informatika
Pengolahan Bahasa Daerah

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment