Skripsi

ANALISIS SENTIMEN PUBLIK TERHADAP CHATGPT DI PLATFORM X MENGGUNAKAN ROBERTA

---

ChatGPT is an artificial intelligence technology that triggers diverse public opinions on social media, particularly on the X platform. The high volume of data and unstructured text characteristics make manual mapping of public perception difficult. Therefore, this study aims to develop an automated sentiment analysis system to classify public opinion toward ChatGPT into Negative, Neutral, and Positive categories using the RoBERTa (A Robustly Optimized BERT Pretraining Approach) model. The dataset used is secondary data totaling 975,450 records, consisting of 209,306 Negative labels, 442,168 Neutral labels, and 323,976 Positive labels. The research stages include text preprocessing, data splitting with an 80:10:10 ratio, and the fine-tuning process of the RoBERTa-base model. Testing was conducted through 6 experimental scenarios with variations in learning rate, batch size, and epoch. The results indicate that Scenario 4, with a learning rate of 2e-5, a batch size of 32, and 3 epochs, is the most optimal configuration for handling the imbalanced dataset. This best-performing model achieved an accuracy of 87.09%, a macro avg f1-score of 86.99%, a macro avg recall of 87.35%, and a macro avg precision of 86.79%.

---

Availability

Detail Information

Series Title

-

Call Number

T1959332026

Publisher

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

Collation

xiv, 134 hlm.; ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

006.312 07

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Analisis sentimen
Prodi Teknik Informatika

Specific Detail Info

-

Statement of Responsibility

TUTI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment