ANALISIS SENTIMEN ULASAN PENGGUNA ALLO BANK MENGGUNAKAN K-NEAREST NEIGHBORS

Authors

  • Marlina STMIK Kharisma Makassar
  • Antonius Indra Dharma Prasetya Universitas Ciputra Surabaya

DOI:

https://doi.org/10.52972/hoaq.vol15no2.p80-87

Keywords:

Analisa sentimen, K-Nearest Neighbors, KNN, Allo Bank, K optimal.

Abstract

Analisis sentimen merupakan teknik penting dalam memahami opini publik terhadap produk atau layanan, termasuk dalam konteks perbankan digital seperti Allo Bank. Sebagai bank digital yang baru beroperasi, Allo Bank perlu memahami persepsi dan kebutuhan pelanggan untuk membangun kepercayaan serta meningkatkan kualitas layanan. Penelitian ini bertujuan untuk mengklasifikasikan ulasan pengguna Allo Bank yang terdapat di Play Store ke dalam kategori sentimen positif, negatif, atau netral, serta untuk menentukan nilai K optimal pada algoritma K-Nearest Neighbors (KNN) melalui berbagai teknik optimasi. Selain itu, kinerja model KNN dievaluasi menggunakan metrik akurasi, presisi, dan recall. Hasil penelitian menunjukkan bahwa 56,4% ulasan bersifat positif, mencerminkan tingkat kepuasan pengguna, sementara 38,8% ulasan negatif mengindikasikan area yang perlu perbaikan, dan hanya 4,8% ulasan yang bersifat netral. Penelitian ini juga menemukan bahwa nilai K optimal dalam algoritma KNN bervariasi tergantung pada teknik optimasi yang digunakan, yaitu cross-validation, grid search, random search, dan elbow method. Variasi ini terjadi karena setiap teknik memiliki tujuan, metrik evaluasi, dan ruang pencarian hyperparameter yang berbeda. Oleh karena itu, perlu dilakukan analisis lebih lanjut untuk menentukan nilai K yang paling sesuai dengan karakteristik data dan tujuan penelitian. Selain itu, hasil analisis menunjukkan bahwa kinerja teknik optimasi sangat dipengaruhi oleh proporsi data latih dan data uji. Teknik cross-validation memberikan hasil terbaik pada proporsi data uji: data latih sebesar 80:20, berdasarkan metrik evaluasi seperti akurasi, presisi, dan recall. Namun, pada proporsi 70:30, tidak ada teknik optimasi yang secara konsisten lebih unggul. Temuan ini menunjukkan bahwa pemilihan teknik optimasi perlu disesuaikan dengan konteks dan karakteristik data untuk memastikan performa model yang optimal.

 

Sentiment analysis is an important technique in understanding public opinion towards products or services, including in the context of digital banking such as Allo Bank. As a newly operating digital bank, Allo Bank needs to understand customer perceptions and needs to build trust and improve service quality. This study aims to classify Allo Bank user reviews on the Play Store into positive, negative, or neutral sentiment categories, and to determine the optimal K value in the K-Nearest Neighbors (KNN) algorithm through various optimization techniques. In addition, the performance of the KNN model was evaluated using accuracy, precision, and recall metrics. The results showed that 56.4% of reviews were positive, reflecting the level of user satisfaction, while 38.8% of negative reviews indicated areas that needed improvement, and only 4.8% of reviews were neutral. This study also found that the optimal K value in the KNN algorithm varies depending on the optimization technique used, namely cross-validation, grid search, random search, and elbow method. This variation occurs because each technique has different objectives, evaluation metrics, and hyperparameter search spaces. Therefore, further analysis is needed to determine the K value that best suits the characteristics of the data and the objectives of the study. In addition, the analysis results show that the performance of optimization techniques is greatly influenced by the proportion of training data and test data. The cross-validation technique gives the best results at a proportion of test data:training data of 80:20, based on evaluation metrics such as accuracy, precision, and recall. However, at a proportion of 70:30, no optimization technique is consistently superior. This finding suggests that the selection of optimization techniques needs to be adjusted to the context and characteristics of the data to ensure optimal model performance.

References

S. Rahayu, Y. Mz, J. Bororing, and R. Hadiyat, "Implementasi Metode K-Nearest Neighbor (K-NN) untuk Analisis Sentimen Kepuasan Pengguna Aplikasi Teknologi Finansial Flip", Edumatic: Jurnal Pendidikan Informatika, vol. 6, no. 1, p. 98-106, 2022. https://doi.org/10.29408/edumatic.v6i1.5433.

A. Putra and S. Juanita, "Analisis Sentimen pada Ulasan Pengguna Aplikasi Bibit dan Bareksa dengan Algoritma KNN", JATISI (Jurnal Teknik Informatika dan Sistem Informasi), vol. 8, no. 2, p. 636-646, 2021. https://doi.org/10.35957/jatisi.v8i2.962.

Ratih Puspitasari, Y. Findawati, and M. Rosid, "Sentiment Analysis of Post-Covid-19 Inflation Based on Twitter Using the K-Nearest Neighbor and Support Vector Machine Classification Methods", Jurnal Teknik Informatika (Jutif), vol. 4, no. 4, p. 669-679, 2023. https://doi.org/10.52436/1.jutif.2023.4.4.801.

A. Maharani and A. Triayudi, "Sentiment Analysis of Indonesian Digital Payment Customer Satisfaction Towards Gopay, Dana, and Shopeepay Using Naïve Bayes And K-Nearest Neighbour Methods", Jurnal Media Informatika Budidarma, vol. 6, no. 1, p. 672, 2022. https://doi.org/10.30865/mib.v6i1.3545.

W. Winarto, I. A. Musdar, and H. Hasniati, "Sentiment analysis of 2024 Presidential Candidate Using the Support Vector Machine Algorithm on Twitter," KHARISMA Tech, vol. 19, no. 1, pp. 86–98, 2024. https://doi.org/10.55645/kharismatech.v19i1.464.

S. Ernawati and R. Wati, "Sentiment Analysis of Three-Period Polemics Using K-Nearest Neighbor With Tf-Idf Weighting", Jurnal Riset Informatika, vol. 4, no. 3, p. 215-220, 2022. https://doi.org/10.34288/jri.v4i3.377A. Sari, T. Hermanto, and M. Defriani, "Sentiment Analysis of Tourist Reviews Using K-Nearest Neighbors Algorithm and Support Vector Machine", Sinkron, vol. 8, no. 3, p. 1366-1378, 2023. https://doi.org/10.33395/sinkron.v8i3.12447.

A. Sari, T. Hermanto, and M. Defriani, "Sentiment Analysis of Tourist Reviews Using K-Nearest Neighbors Algorithm and Support Vector Machine", Sinkron, vol. 8, no. 3, p. 1366-1378, 2023. https://doi.org/10.33395/sinkron.v8i3.12447.

R. Amardita, A. Adiwijaya, and M. Purbolaksono, "Analisis Sentimen Terhadap Ulasan Paris Van Java Resort Lifestyle Place di Kota Bandung Menggunakan Algoritma KNN", JURIKOM (Jurnal Riset Komputer), vol. 9, no. 1, p. 62, 2022. https://doi.org/10.30865/jurikom.v9i1.3793.

I. Iswanto, T. Tulus, and P. Poltak, "Comparison of Feature Selection to Performance Improvement of K-Nearest Neighbor Algorithm in Data Classification", Jurnal Teknik Informatika (Jutif), vol. 3, no. 6, p. 1709-1716, 2022. https://doi.org/10.20884/1.jutif.2022.3.6.471.

P. Rinanda, B. Delvika, S. Nurhidayarnis, N. Abror, and A. Hidayat, "Perbandingan klasifikasi Antara Naive Bayes dan K-Nearest Neighbor Terhadap Resiko Diabetes pada Ibu Hamil", MALCOM: Indonesian Journal of Machine Learning and Computer Science, vol. 2, no. 2, p. 68-75, 2022. https://doi.org/10.57152/malcom.v2i2.432.

W. Ji, "Research on Life Expectancy Prediction Based on Logistic Regression and KNN", BCP Business & Amp; Management, vol. 40, p. 174-183, 2023. https://doi.org/10.54691/bcpbm.v40i.4378.

M. Kabir, M. Islam, A. Kabir, A. Haque, and M. Rhaman, "Detection of Depression Severity Using Bengali Social Media Posts on Mental Health: Study Using Natural Language Processing Techniques", JMIR Formative Research, vol. 6, no. 9, p. e36118, 2022. https://doi.org/10.2196/36118.

S. Williams, S. Bariselli, C. Palego, R. Holland, and P. Coss, "A comparison of Machine-Learning Assisted Optical and Thermal Camera Systems for Beehive Activity Counting", SSRN Electronic Journal, 2021. https://doi.org/10.2139/ssrn.3990570.

M. Özate?, A. Yaman, F. Salami, S. Campos, S. Wolf, and U. Schneider, "Identification and Interpretation of Gait Analysis Features and Foot Conditions by Explainable Ai", Scientific Reports, vol. 14, no. 1, 2024. https://doi.org/10.1038/s41598-024-56656-4.

A. Purpura, D. Giorgianni, G. Orrù, G. Melis, and G. Sartori, "Identifying Single-Item Faked Responses in Personality Tests: a New Tf-Idf-Based Method", Plos One, vol. 17, no. 8, p. e0272970, 2022. https://doi.org/10.1371/journal.pone.0272970.

Downloads

Published

26-12-2024

How to Cite

Marlina, & Prasetya, A. I. D. (2024). ANALISIS SENTIMEN ULASAN PENGGUNA ALLO BANK MENGGUNAKAN K-NEAREST NEIGHBORS. HOAQ (High Education of Organization Archive Quality) : Jurnal Teknologi Informasi, 15(2), 80–87. https://doi.org/10.52972/hoaq.vol15no2.p80-87