Analisis Sentimen Ulasan Aplikasi DLU Ferry Pada Google Play Store Menggunakan Bidirectional Encoder Representations from Transformers

Akhmad, Ekka Pujo Ariesanto

Jurnal Aplikasi Pelayaran dan Kepelabuhanan
Vol 13 No 2 (2023): bulan Maret

DLU Ferry is an application issued by PT. Dharma Lautan Utama to make it easier for customers to order tickets. The DLU Ferry application still has weaknesses, therefore it needs improvement to improve the quality of this application. To find out the weaknesses of this application can be obtained from reviews written by consumers on the Google Play Store. This research contains analysis of consumer review data on the DLU Ferry application. This study uses data from 1575 reviews consisting of positive, neutral and negative sentiments. This research will measure the performance of the Bidirectional Encoder Representations from Transformers (BERT) method in classifying sentiments using the IndoBERT-base Pretrained model with fine-tuning techniques. The test results in this study obtained an accuracy of 86% with the selection of hyperparameters, namely batch size 32, learning rate 3e-6, and epoch 5.