SKRIPSI IF
COMPARISION ANALYSIS SENTIMENT HASTAG KABUR AJA DULU (#KABURAJADULU) ON X PLATFORM BASED ON LOGISTIC REGRESSION ALGORITHM
ABSTRAK
Perkembangan teknologi digital telah mendorong media sosial menjadi ruang publik baru dalam menyuarakan opini masyarakat. Salah satu fenomena menarik yang muncul adalah penggunaan hashtag #KaburAjaDulu di platform X (sebelumnya Twitter), yang merefleksikan reaksi publik terhadap berbagai situasi sosial yang dianggap memicu kekecewaan atau keputusasaan. Penelitian ini bertujuan untuk menganalisis sentimen publik terhadap hashtag tersebut menggunakan pendekatan Natural Language Processing (NLP). Data dikumpulkan dari 6.427 tweet yang mengandung tagar #KaburAjaDulu, kemudian diproses melalui tahapan preprocessing, pelabelan sentimen berbasis lexicon, serta klasifikasi menggunakan algoritma Logistic regression, Random Forest, dan Support Vector Machine (SVM). Untuk mengatasi ketidakseimbangan kelas data, diterapkan metode SMOTE (Synthetic Minority Over-sampling Technique). Hasil penelitian menunjukkan bahwa penerapan NLP mampu mengidentifikasi sentimen publik dengan akurasi yang meningkat secara signifikan setelah penerapan SMOTE. Model Logistic regression menunjukkan performa terbaik dengan akurasi 84%. Mayoritas sentimen publik terhadap hashtag ini bersifat netral, disusul sentimen negatif dan positif. Penelitian ini memberikan kontribusi dalam pemetaan opini publik digital serta menjadi contoh implementasi NLP untuk analisis isu sosial di media sosial berbahasa Indonesia.
Kata Kunci: Analisis Sentimen, Natural Language Processing (NLP), Twitter, #KaburAjaDulu, Lexicon-Based, SMOTE, Klasifikasi Teks
ABSTRACT
The advancement of digital technology has transformed social media into a new public space for expressing opinions. One notable phenomenon is the emergence of the hashtag #KaburAjaDulu on the X platform (formerly Twitter), which reflects public reactions to various social situations that trigger disappointment or despair. This study aims to analyze public sentiment toward the hashtag using a Natural Language Processing (NLP) approach. The dataset comprises 6,427 tweets containing the hashtag #KaburAjaDulu, which were processed through several stages including preprocessing, sentiment labeling using a lexicon-based method, and classification using Logistic regression, Random Forest, and Support Vector Machine (SVM) algorithms. To address class imbalance in the dataset, the SMOTE (Synthetic Minority Over-sampling Technique) method was applied. The results show that NLP effectively identifies public sentiment, with model performance improving significantly after applying SMOTE. Logistic regression achieved the highest accuracy at 84%. The majority of public sentiment toward the hashtag is neutral, followed by negative and positive sentiments. This study contributes to understanding public digital opinion trends and provides a practical implementation of NLP for analyzing social issues in Indonesian-language social media.
Keywords: Sentiment Analysis, Natural Language Processing (NLP), Twitter, #KaburAjaDulu, Lexicon-Based, SMOTE, Text Classification
| S251002012 | 607.2 NUG c | Perpustakaan STMIK AMIKBANDUNG | Tersedia |
Tidak tersedia versi lain