Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Comparison Performance of SVM, Naïve Bayes and XGBoost Classifier on Climate Change Issue

Rektorat UDINUS 2024 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2024
Goals
Record
506688ab-e81f-40ec-860f-a1dd20a53e1b
Published
2026-10-30

Universitas Dian Nuswantoro (UDINUS), through researchers Anhsori Khusman and Guruh Fajar Shidik, conducted a study examining attitudes toward the climate change issue through sentiment analysis and text classification. Climate change is described as a long-term shift in weather patterns from tropical to polar regions. The source identifies it as a worldwide threat that is beginning to materialize and exert pressure on numerous industries. The study examined public opinion and perceived issue risk using information shared on Twitter, a microblogging platform where thousands of people share thoughts every day.

The activity used a dataset obtained from Kaggle. The dataset was divided into four sentiment categories: News, Pro, Neutral, and Anti. It was also divided into 80% training data and 20% testing data. UDINUS researchers used three classifier algorithms to examine attitudes about the climate change issue: Support Vector Machine (SVM), Naïve Bayes (NB), and XGBoost (XGB). These methods were considered within the context of sentiment analysis in natural language processing and text classification.

In the classification results, the SVM classifier achieved an accuracy of 73.32% and beat the NB and XGB classifiers. Hyperparameter optimization was used to optimize accuracy, and the result increased by 3% in the SVM method. The activity therefore presents a comparison of three classifier algorithms for examining attitudes about climate change in data obtained from Kaggle and related to public opinion expressed through Twitter. The findings identify SVM as the classifier with the highest reported accuracy among the methods examined. Through this research activity, UDINUS addresses climate change as a subject of public-opinion analysis and algorithmic classification.

Reference https://www.scopus.com/pages/publications/85213304695?origin=resultslist