@misc{eprints75255,
           month = {Agustus},
           title = {ANALISIS SENTIMEN TERHADAP KINERJA PELAYANAN
DI PT BANK RAKYAT INDONESIA (PERSERO) TBK.
MENGGUNAKAN METODE SUPPORT VECTOR MACHINE,
NAIVE BAYES, DAN K-NEAREST NEIGHBORS},
          author = {RAMADHIA FITRI  DEVI },
         address = {UNIVERSITAS LAMPUNG},
       publisher = {FAKULTAS MATEMATIKA DAN ILMU PENGETAHUAN ALAM},
            year = {2023},
             url = {http://digilib.unila.ac.id/75255/},
        abstract = {Penelitian ini mengenai analisis sentimen terhadap kinerja pelayanan di PT. Bank
Rakyat Indonesia (Persero) Tbk. Dataset yang digunakan sebanyak 1652 data Xs
yang dikumpulkan sejak bulan Maret 2022 sampai dengan Februari 2023 yang
diklasifikasikan menjadi dua kelas, yaitu sentimen positif dan sentimen negatif.
Dalam memberikan pelayanan kepada nasabah antrian yang panjang kerap terjadi
PT. Bank Rakyat Indonesia (Persero) Tbk. sehingga dapat menyebabkan komentarkomentar
mengenai pelayanan. PT. Bank Rakyat Indonesia (Persero) Tbk. belum
terdapat dalam menanggapi komentar nasabah seharusnya dibuatkan secara rutin
analisis sentimen terhadap kinerja pelayanan. Analisis sentimen dalam penelitian
ini menggunakan metode Support Vector Machine (SVM), Na{\"i}ve Bayes, dan KNearest
Neighbors. Penelitian ini menerapkan penyeimbangan data pada dataset
inbalanced menggunakan Synthetic Minority Over-sampling Technique (SMOTE)
untuk mendapatkan performa yang terbaik. Hasil evaluasi menggunakan algoritma
Support Vector Machine (SVM) pada dataset balanced yang menghasilkan kinerja
yang lebih baik dibandingkan dengan metode lainnya akurasi sebesar 94,30\%,
algortime Na{\"i}ve Bayes sebesar 93,68\% pada dataset balanced, sementara algoritma
K-Nearest Neighbors (KNN) sebesar 88,31\% pada dataset balanced. Perbandingan
hasil kinerja berdasarkan ketiga metode yang telah digunakan menunjukan bahwa
algoritma Support Vector Machine (SVM) lebih baik dibandingkan Na{\"i}ve Bayes,
dan K-Nearest Neighbors (KNN) untuk analisis sentimen terhadap kinerja
pelayanan di PT. Bank Rakyat Indonesia (Persero) Tbk.
Kata Kunci: Analisis Sentimen, Support Vector Machine, Na{\"i}ve Bayes, K-Nearest
Neighbors, Pelayanan PT. Bank Rakyat Indonesia (Persero) Tbk.

This research is about sentiment analysis of the service performance at PT. Bank
Rakyat Indonesia (Persero) Tbk. The dataset used is 1652 data Xs collected from
March 2022 to February 2023, classified into two classes, namely positive and
negative sentiment. In providing services to customers, long queues often occur,
PT. Bank Rakyat Indonesia (Persero) Tbk. so that it can lead to comments about
the service. PT. Bank Rakyat Indonesia (Persero) Tbk. Yet to be available in
response to customer comments, we should routinely make sentiment analysis on
service performance. Sentiment analysis in this study uses the Support Vector
Machine (SVM), Na{\"i}ve Bayes, and K-Nearest Neighbors methods. This study
applies data balancing to inbalanced datasets using the Synthetic Minority Oversampling
Technique (SMOTE) for the best performance. The evaluation results use
the Support Vector Machine (SVM) algorithm on a balanced dataset, producing
better performance than other methods with an accuracy of 94,30\%. The Na{\"i}ve
Bayes algorithm is 93,68\% on a balanced dataset, while the K-Nearest Neighbors
(KNN) algorithm is 88,31\%. The performance comparison results based on the
three methods used show that the Support Vector Machine (SVM) algorithm is
better than the Na{\"i}ve Bayes and K-Nearest Neighbors (KNN) algorithms for
sentiment analysis of service performance at PT. Bank Rakyat Indonesia (Persero)
Tbk.
Keywords: Sentiment Analysis, Support Vector Machine, Na{\"i}ve Bayes, K-Nearest
Neighbors, service of PT. Bank Rakyat Indonesia (Persero) Tbk.}
}