Nugroho, Alif Suryo and ', Widi Widayat, S.Kom., M.Eng. (2025) Perbandingan Performa Algoritma Machine Learning Untuk Analsis Sentimen Pengguna X Mengenai Topik Deepseek Al. Skripsi thesis, Universitas Muhammadiyah Surakarta.
![[img]](https://eprints.ums.ac.id/style/images/fileicons/application_pdf.png) |
PDF (Surat Pernyataan Publikasi)
UNGGAH Suryo.pdf Download (306kB) |
|
PDF (Naskah Publikasi)
SKRIPSI-DONE_L200200004_ALIF_SURYO_NUGROHO.pdf Download (663kB) |
Abstract
This study analyzes user sentiment on X (Twitter) regarding the launch of DeepSeek AI using five classification algorithms: Naïve Bayes, Random Forest, Logistic Regression, KNN, and SVM. Data was collected via Tweet Harvest and labeled using both automatic (TextBlob) and manual methods. To validate manual labeling consistency, the Cohen’s Kappa method was applied. The SMOTE technique was used to address class imbalance. Each algorithm was tested across five tuning versions. Evaluation results show that SVM with V4 (aggressive tuning) achieved the best performance, with an F1-Score of 85.86%, accuracy of 85.98%, precision of 86.06%, and recall of 85.98%. These findings indicate that SVM is highly effective for sentiment analysis on complex social media data.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Uncontrolled Keywords: | AI, Machine Learning, Sentiment Analysis, Twitter. |
| Subjects: | T Technology > T Technology (General) T Technology > TZ Technical Information |
| Divisions: | Fakultas Komunikasi dan Informatika > S1 Teknik Informatika |
| Depositing User: | ALIF SURYO NUGROHO |
| Date Deposited: | 22 Jul 2025 06:50 |
| Last Modified: | 22 Jul 2025 06:50 |
| URI: | http://eprints.ums.ac.id/id/eprint/136237 |
Actions (login required)
 |
View Item |
