Sentiment Analysis Of Public Figure News Using Tree-Based Algorithm

Karimah, Amelia Mumtazah (2025) Sentiment Analysis Of Public Figure News Using Tree-Based Algorithm. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5025201128-Undergraduate-Thesis.pdf Restricted to Repository staff only Download (17MB) | Request a copy

Abstract

Tokoh Publik seperti politisi, selebriti, dan pemimpin bisnis sering kali menjadi bahan pemberitaan di media daring. Pemberitaan ini bisa membentuk pandangan masyarakat dan memengaruhi citra mereka secara langsung. Berdasarkan data APJII tahun 2022–2023, jumlah pengguna internet di Indonesia mencapai lebih dari 215 juta orang, dan 84% di antaranya mengakses berita melalui media online seperti Detikcom. Melihat besarnya pengaruh pemberitaan digital, penting untuk mengetahui bagaimana sentimen publik terhadap berita yang berkaitan dengan figur publik. Penelitian ini bertujuan membangun model klasifikasi sentimen menggunakan algoritma berbasis pohon seperti Decision Tree, Random Forest, dan XGBoost, serta tiga metode ekstraksi fitur: TF-IDF, CBOW, dan Skip-gram. Data diperoleh melalui web scraping dan dikelompokkan menjadi sentimen positif, netral, dan negatif. Evaluasi dilakukan dengan beberapa pendekatan, termasuk pembagian data (80:20), K-Fold cross-validation, dan SMOTE untuk menangani data yang tidak seimbang. Hasil penelitian menunjukkan bahwa performa model bergantung pada kombinasi metode yang digunakan. XGBoost dengan Skip-gram menghasilkan akurasi tertinggi sebesar 71,19%. Di sisi lain, Random Forest yang dipadukan dengan TF-IDF dan SMOTE memberikan hasil yang stabil dan cukup kuat, dengan akurasi mencapai 65,54% serta nilai presisi, recall, dan F1-score yang seimbang. Sementara itu, Decision Tree menunjukkan performa paling rendah, terutama saat berhadapan dengan data yang tidak seimbang. Model terbaik dari penelitian ini kemudian diimplementasikan ke dalam antarmuka pengguna sederhana yang dapat melakukan klasifikasi sentimen secara langsung. Temuan ini menunjukkan bahwa algoritma berbasis pohon, khususnya XGBoost dan Random Forest, sangat potensial untuk digunakan dalam analisis sentimen berita mengenai figur public
=======================================================================================================================================
Public figures like politicians, celebrities, and business leaders are frequently featured in online news, and how they are portrayed can strongly influence public perception. With over 215 million internet users in Indonesia and 84% of them relying on online media for news, sentiment analysis has become an important tool for understanding how the public reacts to news involving these figures. This study aims to build a sentiment classification model using three tree-based algorithms Decision Tree, Random Forest, and XGBoost along with three feature extraction methods: TF-IDF, CBOW, and Skip-gram. The dataset was collected through web scraping and categorized into positive, neutral, and negative sentiments. Several evaluation techniques were applied, including train-test split, K-Fold cross-validation, and SMOTE to handle imbalanced data. The results show that model performance varied depending on the combination of algorithm and feature extraction method. XGBoost with Skip-gram achieved the highest accuracy at 71.19%, while Random Forest with TF-IDF and SMOTE showed strong and consistent results, reaching 65.54% accuracy with balanced precision, recall, and F1-score. Decision Tree, although simple and interpretable, performed the weakest, especially with imbalanced data. The most effective models were then integrated into a simple user interface that can perform real-time sentiment classification. Overall, the study highlights that tree-based algorithms particularly XGBoost and Random Forest are effective tools for analyzing public sentiment in online news about public figures.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Analisis Sentimen, Berita, Decision Tree, Random Forest, XGBoost, Decision Tree, News, Random Forest, Sentiment Analysis, XGBoost
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.62 Decision support systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Amelia Mumtazah Karimah
Date Deposited:	28 Jul 2025 10:28
Last Modified:	28 Jul 2025 10:28
URI:	http://repository.its.ac.id/id/eprint/122205

Actions (login required)

View Item