Fadhilah, Nur Anisah and Asri, Yessy and Palupiningsih, Pritasari (2025) PENERAPAN NAMED ENTITY RECOGNITION (NER) MENGGUNAKAN METODE RULE-BASED PADA ULASAN PENGADUAN DAN PELAYANAN PLN MOBILE. Diploma thesis, ITPLN.
202131020_Nur Anisah Fadhilah_Revisi_Skripsi_Nur Anisah Fadhilah.pdf
Restricted to Registered users only
Download (7MB)
Abstract
Aplikasi PLN Mobile menjadi sarana utama bagi pelanggan dalam menyampaikan pengaduan dan mendapatkan layanan terkait kelistrikan. Penelitian ini bertujuan untuk menerapkan Named Entity Recognition (NER) menggunkan metode rule-based guna pelabelan entitas penting dalam ulasan pengaduan dan pelayanan. Metode rule-based digunakan untuk membuat pola aturan dalam bentuk keyword yaitu O2O, Feature Error, Umum, Feature Design, dan Registration agar dapat dilabelkan oleh NER berdasarkan pola aturan yang telah dibuat. Selain itu, penelitian ini mengevaluasi tingkat akurasi pelabelan entitas menggunakan arsitektur Bidirectional Long Short-Term Memory (BiLSTM). Hasil penelitian menunjukkan bahwa pendekatan rule-based dapat mengenali entitas dengan baik meskipun terdapat variasi bahasa dalam ulasan. Sementara itu, model BiLSTM menghasilkan akurasi 1.00 dengan precision, recall, dan F1-score yang sempurna pada sebagian besar kelas, menunjukkan keandalannya dalam pelabelan entitas pada teks ulasan PLN Mobile.
The PLN Mobile application serves as the primary platform for customers to submit complaints and receive electricity-related services. This study aims to implement rule-based Named Entity Recognition (NER) to identify and classify key entities in complaint and service reviews. The rule-based method is employed to detect entities such as names, locations, complaint types, and other relevant information in user reviews. Additionally, this study evaluates the accuracy of entity labeling using the Bidirectional Long Short-Term Memory (BiLSTM) architecture. The results show that the rule-based approach effectively identifies entities despite language variations in the reviews. Meanwhile, the BiLSTM model achieves a perfect accuracy of 1.00, with optimal precision, recall, and F1-score across most classes, demonstrating its reliability in entity classification for PLN Mobile reviews.
| Item Type: | Thesis (Diploma) |
|---|---|
| Uncontrolled Keywords: | PLN Mobile, Named Entity Recognition, rule-based, Bi-LSTM PLN Mobile, Named Entity Recognition, rule-based, Bi-LSTM |
| Subjects: | Skripsi Bidang Keilmuan > Teknik Informatika |
| Divisions: | Fakultas Telematika Energi > S1 Teknik Informatika |
| Depositing User: | Sudarman |
| Date Deposited: | 09 Oct 2025 05:54 |
| Last Modified: | 09 Oct 2025 05:54 |
| URI: | https://repository.itpln.ac.id/id/eprint/1987 |
