Comparative Analysis of Sentiment Analysis Using the Support Vector Machine and Naive Bayes Algorithm on Cryptocurrencies

Abstract

Objective – Cryptocurrency is growing overtime even being adopted as a legal money in a country out there. Besides can be used as a money, cryptocurrency also can be used as a digital goods to be trade and investment assets. To do some investing in cryptocurrency, there’s a need to evaluate the fundamental and sentiment of that cryptocurrency. This study aims to evaluate cryptocurrency based on responses of Twitter user.Methodology – The Algorithms used in this sentiment analysis study are Support Vector Machine and Naïve Bayes because it’s already proven that these 2 algorithm able to give a good accuracy and performance and using CRISP – DM framework for the study flow.Findings – This research predicts the sentiment for Bitcoin, Ethereum, Binance Coin, Dogecoin, and Ripple using the CRISP - DM method and using Support Vector Machine and Naïve Bayes algorithm.Novelty – This study calculate the sentiment on cryptocurrency using Rapidminer tools.Limitations - This study uses Bitcoin, Ethereum, Binance Coin, Dogecoin, and Ripple using tools such as rapidminerKeywords — Cryptocurrency, Naïve Bayes, Sentiment Analysis, Support Vector Machine