Skripsi/Tugas Akhir
Identifikasi Emosional pada Sentimen Negatif Pengaduan MyIndiHome Menggunakan Metode Support Vector Machine
ABSTRAK
MyIndiHome, layanan internet dan televisi kabel dari Grapari Telkomsel, menjadi pilihan utama di Indonesia namun sering dikritik pengguna terkait masalah teknis seperti koneksi yang tidak stabil, kesulitan login, dan harga mahal. Identifikasi sentimen terhadap 500 komentar negatif yang merujuk pada aplikasi play store dari Mei hingga Juni 2024 menggunakan Support Vector Machine (SVM) menunjukkan akurasi 93%, dengan precision 87,5%, recall 100%, dan f1-score 93,33%, untuk mengklasifikasikan emosi marah dan santai. Scatter plot hyperplane memisahkan kelas sentimen, dan wordcloud menyoroti kata-kata seperti "sampah" dan "jelek" sebagai tema dominan dalam keluhan marah. Dengan teknik TF-IDF, SVM mampu mengidentifikasi sumber utama ketidakpuasan pengguna, membantu Grapari Telkomsel mengambil langkah perbaikan. Implementasi algoritma ini dilakukan di Google Colab untuk mempercepat proses pelatihan. Identifikasi ini berperan penting dalam meningkatkan kualitas layanan, kepuasan, dan loyalitas pelanggan MyIndiHome serta menjaga reputasi layanan tersebut.
Kata kunci: Grapari Telkomsel, MyIndiHome, Play Store, Sentimen Negatif, Emosional, Support Vector Machine (SVM), Identifikasi
ABSTRACT
MyIndiHome, an internet and cable television service from Grapari Telkomsel, has become a top choice in Indonesia, though it often receives criticism from users regarding technical issues such as unstable connections, login difficulties, and high prices. Sentiment analysis of 500 negative comments from the Play Store, dated between May and June 2024, using a Support Vector Machine (SVM) model, showed an accuracy of 93%, with a precision of 87.5%, recall of 100%, and an F1- score of 93.33% for classifying sentiments of anger and calmness. A scatter plot with a hyperplane separates sentiment classes, while a word cloud highlights keywords like “trash” and “bad” as dominant themes in angry complaints. Using the TF-IDF technique, SVM effectively identifies the primary sources of user dissatisfaction, helping Grapari Telkomsel take corrective actions. The algorithm was implemented in Google Colab to expedite the training process. This identification plays a crucial role in improving service quality, customer satisfaction, and loyalty to MyIndiHome, as well as maintaining the service’s reputation.
Keywords: Grapari Telkomsel, MyIndiHome, Play Store, Negative Sentiment, Emotional, Support Vector Machine (SVM), Identification
Tidak ada salinan data
Universitas DIPA Makassar
NPP 7371142D1000002
Jln. Perintis Kemerdekaan KM.9
Telp. (0411)587194
Hotline: +6281228221994
WhatsApp Admin: +6281342092072
e-Mail: [email protected]
© 2024 — Perpustakaan UNDIPA Makassar - SLiMS