Analisis Sentimen Mobil Listrik di Indonesia Menggunakan Long-Short Term Memory (LSTM)

Abstract

Vehicles using fuel oil that is converted into mechanical energy were introduced in 1891 by John W. Lambert in America. But with this, the level of air pollution caused by exhaust emissions has become a problem today, until environmentally friendly engine innovations appear. The beginning of the development of these innovations was marked by hybrid technology cars. This technology has not completely abandoned the use of oil as fuel. In general, these vehicles are known as HEV or Hybrid Electic Vehicles. Then came a car that was entirely with an electric motor drive or EV or Electric Vehicle. Although the technology is considered environmentally friendly, on the other hand it does not make all elements of society accept any changes, especially in fuel oil engines to electric motors. With these changes, there are pros and cons that are the focus of researchers by utilizing sentiment analysis which is a Natural Language Processing (NLP) scientific family to analyze what aspects make society pro or con to the emergence of environmentally friendly vehicles. Data collection in this study took from YouTube comments in the form of Indonesian text data carried out using Python programming language and Long-Short Term Memory (LSTM) as an algorithm for analyzing public opinion. The dataset was divided into training data and test data with a ratio of 67:33, The results showed that the model can be used on Indonesian text data well. Then for the process of accuracy test data 63%, then macro avg precision 62%, macro avg recall 60%, macro avg f1-score 60%, weighted avg precision 62%, weighted avg recall 63%, weighted avg f1-score 62%, roc_auc 81%. In this study, it can also be seen that the topic of discussion that often arises, namely prices in all classes. However, negative sentiment is more than other sentiment classes, one of which is due to electric car manufacturers so it is very necessary to pay attention to stakeholders regarding prices that are suitable for the Indonesian market.