Skripsi

DETEKSI HUMOR TEKS PENDEK MENGGUNAKAN MODEL INDOBERT

---

The rapid growth of Indonesian text content on social media has increased the need for automatic systems capable of accurately detecting humor, as humor often contains implicit meanings, wordplay, and cultural context. This study develops an Indonesian short-text humor detection system using a fine-tuning approach on the IndoBERT model. The objective of this research is to classify text into two categories, namely humor and non-humor, using binary classification. The dataset consists of Indonesian short texts that have undergone preprocessing and normalization. The fine-tuning process evaluates 12 experimental model configurations by varying maximum sequence length, learning rate, and dropout rate. Model performance is evaluated using accuracy, precision, recall, F1-score, and confusion matrix. The experimental results show that the fine-tuned IndoBERT model achieves strong performance, with the best configuration obtaining 97.55% accuracy, 97.78% precision, 97.70% recall, and a 97.51% F1-score. Further analysis indicates that configurations using a maximum sequence length of 256 and learning rates in the range of 1e-5 to 2e-5 provide the most optimal performance, particularly based on the F1-score metric, which reflects a balanced trade-off between precision and recall. The trained models are implemented into a web-based application using Streamlit, providing single text analysis and model performance statistics. These results indicate that IndoBERT fine-tuning is highly effective for Indonesian humor detection and suitable for practical deployment. Keywords: humor detection, text classification, IndoBERT, fine-tuning, F1-score

---

Availability

Detail Information

Series Title

-

Call Number

T1900562025

Publisher

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

Collation

xv, 158 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

005.107

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Teknik Informatika
Pengembangan Software

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment