Analisis Sentimen pada Review Konsumen Menggunakan Metode Naive Bayes dengan Seleksi Fitur Chi Square untuk Rekomendasi Lokasi Makanan Tradisional

Novan Dimas Pratama (Fakultas Ilmu Komputer, Universitas Brawijaya)
Yuita Arum Sari (Fakultas Ilmu Komputer, Universitas Brawijaya)
Putra Pandu Adikara (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 2 No 9 (2018): September 2018

Consumer reviews at a restaurant are very influential in the quality of the restaurant itself. Many of the consumers pour critics or opinions through the internet media. The purpose of this study was to analyze the opinion sentiment from traditional food consumers as well as provide location recommendations with the desired keywords. Naive Bayes is a machine learning technique that is often used to classify text data. Chi Square is a feature selection used to calculate the level of a feature's dependencies on a class. In this study, Chi Square method gives value to the feature which is then sorted and selected according to percentage tested. Selected features are used for the classification process using the Naive Bayes method. The result of classification accuracy with 25% feature selection is 81%, with 50% feature selection is 80% and with 77% feature selection is 80%. From this test it can be concluded that feature selection is not so influential on the result value accuracy. It can be seen the difference of the accuracy value between using feature selection and without using a feature selection that is not very significant.