Skripsi

ANALISIS SENTIMEN TERKAIT APLIKASI CHATGPT MENGGUNAKAN BERT DAN RNN

---

Penilaian

---

ChatGPT has become one of the most popular artificial intelligence tools worldwide, but its rapid adoption has also generated diverse user reactions and reviews. Understanding these perceptions is important for evaluating and improving service quality. Sentiment analysis is a suitable approach to explore such opinions. This study employs BERT (Bidirectional Encoder Representations from Transformers) as the embedding technique and RNN (Recurrent Neural Network) as the classification model. Two classification scenarios are used: binary (positive and negative) and multi-class (positive, neutral, and negative). To overcome the limitation of labeled data, a semi supervised pseudo-labeling strategy is applied during fine tuning. Experimental results show that binary classification achieved the highest accuracy of 91%, while the three-label scenario reached only 77%. The lower result in the three label setting is mainly caused by difficulties in identifying ambiguous neutral sentiments and the influence of imbalanced labels. Overall, combining BERT and RNN is effective for sentiment analysis, particularly in binary classification, though challenges remain in multi-label contexts.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2507004381	T179852	T1798522025	Central Library (Reference)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T1798522025

Publisher

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

Collation

xvi, 98 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