International Journal of Applied Information Systems and Informatics
Vol 1, No 1 (2023): November 2023

Analysis Of Twitter User Sentiment To Tiktok Shop Using Naïve Bayes And Decision Tree Algorithms

Jafar Sidiq, Soleh (Unknown)
Nur Rachman, Andi (Unknown)



Article Info

Publish Date
21 Nov 2023

Abstract

The growth of internet users is fantastic, before the pandemic the figure was only 175 million. While the latest data from the Asosiasi Penyelenggaraan Jasa Internet Indonesia (APJII), in 2022 internet users in Indonesia will reach around 210 million. One of the influences on the increasing number of internet users in Indonesia is the increasing number of buying and selling activities through internet media. At this time there are various kinds of e-commerce applications. One of the latest e-commerce in Indonesia is Tiktok Shop. Tiktok shop is a new feature of the Tiktok application which was established on April 17, 2021. The development of Tiktok shop cannot be separated from the people who use this feature. Many people give opinions about Tiktok Shop on one of the social media, namely Twitter. Twitter is a place to get data expressed by the public through tweets posted to the timeline. The data used are tweets in Indonesian with a dataset of 1000 tweets. The data is then processed to be analyzed for knowledge. The analysis is done with Naïve Bayes and Decision Tree methods. The accuracy results of the Naïve Bayes algorithm are 90% and the Decision tree algorithm is 93%, so the Decision Tree algorithm is better for classifying sentiment analysis of twitter users towards Tiktok Shop with a data division of 90%. 

Copyrights © 2023






Journal Info

Abbrev

jaisi

Publisher

Subject

Computer Science & IT

Description

International Journal of Applied Information Systems and Informatics (JAISI) is an international journal in Information Systems and Informatics as a forum for research publications that are systematically arranged based on clear background, appropriate methods, well presented research results, as ...