No abstract

Abstract. Sentiment Analysis Using Naive Bayes Classifier Against Restaurant Reviews in Singapore. Various restaurant options bring up a problem for diners to pick a restaurant to dine in. Thus, visitors usually perceive the restaurant's recommendation or rating in advance to know other diners' opinions about the restaurant. Previous restaurant diners' comments can be presented in sentiment analysis to determine their satisfaction. This research investigates the Naïve Bayes Classifier algorithm's performance in classifying visitors' sentiment based on restaurant diner comments. We will group visitors' comments into two types of sentiment: positive (satisfied) and negative (unsatisfied). The results of the data classification test are analyzed to determine its accuracy. The grouping of visitor satisfaction reviews using the naïve bayes algorithm provides an accuracy score of 73%. Besides, we visualize the research classification results in the browser-based R Shiny web application through word cloud and diagrams.Keywords:restaurant review, sentiment analysis, Naïve Bayes ClassifierAbstrak. Variasi pilihan restoran yang tidak sedikit menjadi salah satu masalah bagi pengunjung ketika ingin memilih restoran. Sehingga, pengunjung biasanya melihat rekomendasi atau penilaian pengunjung lain terhadap restoran tersebut terlebih dahulu untuk mengetahui penilaian pengunjung lain terhadap restoran tersebut. Penilaian atau review pengunjung dapat disajikan dalam analisis sentimen berdasarkan komentar para pengunjung restoran sebelumnya untuk melihat kepuasan pengunjung terhadap restoran tersebut. Penelitian ini dilakukan untuk mengetahui performa algoritma Naïve Bayes Classifier dalam melakukan klasifikasi sentimen berdasarkan komentar pengunjung restoran. Penelitian dilakukan dengan mengklasifikasikan data komentar pengunjung restoran menjadi dua kategori sentimen, yaitu: positif (satisfied) dan negatif (unsatisfied). Hasil pengujian pengklasifikasian data kemudian dianalisis akurasinya. Hasil pengelompokan review kepuasan pengunjung menggunakan algoritma naïve bayes memberikan nilai akurasi sebesar 73%. Visualisasi hasil klasifikasi dari analisis kemudian ditampilkan pada aplikasi berbasis web yaitu R Shiny berupa wordcloud dan diagram. Kata Kunci: penilaian restoran, analisis sentimen, Naïve Bayes Classifier

No abstract

No abstract

This research was conducted to find the groups of elementary schools in the Special Capital Region of Jakarta, also known as DKI Jakarta. Elementary school data were selected because it is the first stage of formal education in Indonesia. This research used K-means clustering with the elbow method to determine optimal cluster numbers. The optimal cluster number is three  with Cluster 2 having the most members, followed by Cluster 1 and Cluster 0. The data distribution of Cluster 2 shows that the second-most student body and public schools located in East and West Jakarta have an adequate student-to-teacher ratio based on Article 17 of Government Regulation 74, 2008.