Analisis Sentemen Terhadap Aplikasi Bukalapak Sebelum IPO dan Sesudah IPO Menggunakan Algoritma Naive Bayes

Bayu Yanuargi

JNANALOKA
Vol. 03 No. 01 Maret Tahun 2022

Bukalapak is one of the earliest eCommerce startups established in Indonesia. Bukalapak has been bridging between sellers (Pelapak) and buyers since 2010. In 2021 Bukalapak ventured to conduct Initial Public Offers on the IDX. There are many kinds of responses from Bukalapak users to Bukalapak's steps, both positive and negative. These negative or positive sentiments can be used as input and evaluation for Bukalapak itself to maintain the loyalty of its users. The research process starts from collecting data obtained from scrapping data on Bukalapak product reviews on Google Playstore before and after the IPO. Then preprocessing the data starting from casefolding, removing stop words, tokenization, steming to TF-IDF. The results of the preprocessing are then used as data for classification using Naive Bayes. The classification was then tested and obtained an accuracy value for the data before the IPO of 77% and the data after the IPO of 76%.