Skripsi

KLASIFIKASI JUDUL BERITA PALSU MENGGUNAKAN METODE BIDIRECTIONAL ENCODER REPRESENTATIONS FROM TRANSFORMERS (BERT)

---

Penilaian

---

This research aims to classify fake news headlines in Indonesian as an initial step in detecting and reducing the spread of misinformation. Headlines are often the first element read and have strong potential to shape public opinion, especially when they contain negative or sensational narratives. Unlike previous studies that analyze full news content or focus only on political news, this study uses the Bidirectional Encoder Representations from Transformers (BERT) approach to analyze news headlines. The dataset consists of 14,231 headlines from various categories, divided into training (6,972), validation (2,989), and test (4,270) sets. Experiments were conducted using nine scenarios that combine different parameters: epochs (10, 20, 30), learning rates (1e-4, 3e-5, 5e-5), and numbers of frozen BERT layers (6, 8, 10). Model performance was evaluated using a confusion matrix and metrics such as accuracy, precision, recall, and F1-score. The best result was achieved with 20 epochs, a learning rate of 3e-5, and 6 frozen layers, reaching an accuracy of 92.01%.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2507002524	T172360	T1723602025	Central Library (Reference)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T1723602025

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xiii, 120 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

004.07

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Teknik Informatika

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment