Skripsi

PENERAPAN KOMBINASI AUGMENTASI EDA, BACK TRANSLATION, DAN BERT UNTUK MENINGKATKAN KINERJA ANALISIS SENTIMEN MENGGUNAKAN GRU PADA ULASAN LAYANAN TELEMEDICINE

---

User reviews of telemedicine services contain important information reflecting user satisfaction and perceptions, which are categorized into three labels: negative, neutral, and positive. Automated sentiment analysis can help categorize reviews to evaluate service quality more efficiently. Gated Reccurent Unit (GRU) requires a large dataset so that the model can optimally recognize patterns. This study applies a combination of Easy Data Augmentation (EDA) augmentation, Back Translation, and Bidirectional Encoder from Transformer (BERT) to address the issue of data scarcity and analyze its impact on GRU performance in sentiment analysis. Applying the combination of EDA augmentation, Back Translation, and BERT to Alodokter review data yielded a BERTScore of 0.8936 and a Distinct-10 score of 0.9998, while on Klikdokter review data, it yielded a BERTScore of 0.6433 and a Distinct-10 of 0.9167, indicating that the Alodokter review data results exhibit higher alignment with the original sentences and excellent diversity. Applying the GRU architecture to Alodokter review data for each label resulted in an average increase in accuracy of 0.25%, precision of 21.09%, recall of 24.58%, and an F1-score of 22.89%, while for Klikdokter review data, accuracy increased by 1.91%, precision of 21.26%, recall of 29.45%, and an F1-score of 24.93%, indicating that the GRU’s performance improved in sentiment analysis, both in terms of prediction accuracy and balance in identifying each label. Based on these results, it can be concluded that the combination of EDA augmentation, back translation, and BERT is effective in improving data quality in terms of semantic similarity and text diversity as well as enhancing the performance of the GRU architecture in sentiment analysis of telemedicine service reviews.

---

Availability

Detail Information

Series Title

-

Call Number

T1982322026

Publisher

Indralaya : Prodi Ilmu Matematika, Fakultas Matematika Dan Ilmu Pengetahuan Alam Universitas Sriwijaya., 2026

Collation

xx, 116 hlm.; ilus.; tab, 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

006.307

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Kecerdasan Buatan
Prodi Ilmu Matematika

Specific Detail Info

-

Statement of Responsibility

TUTI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment