Zahra Agusfiyanti Nurlaila, - (2023) ALGORITMA SUPPORT VECTOR MACHINE (SVM) DAN NAÏVE BAYES PADA ANALISIS SENTIMEN SOSIAL MEDIA TWITTER: Studi Kasus Kuliah Luring Setelah Pandemi Covid-19. S1 thesis, Universitas Pendidikan Indonesia.
![[img]](http://repository.upi.edu/style/images/fileicons/text.png) |
Text
S_MAT_1905657_Title.pdf Download (846kB) |
|
Text
S_MAT_1905657_Chapter1.pdf Download (315kB) |
|
Text
S_MAT_1905657_Chapter2.pdf Restricted to Staf Perpustakaan Download (567kB) |
|
Text
S_MAT_1905657_Chapter3.pdf Download (995kB) |
|
Text
S_MAT_1905657_Chapter4.pdf Restricted to Staf Perpustakaan Download (1MB) |
|
Text
S_MAT_1905657_Chapter5.pdf Download (299kB) |
|
Text
S_MAT_1905657_Appendix.pdf Restricted to Staf Perpustakaan Download (759kB) |
Abstract
Setelah melewati masa pandemi Covid-19, berbagai sektor dalam kehidupan manusia banyak mengalami perubahan, termasuk dalam kegiatan pendidikan. Hal tersebut membuat berbagai pendapat atau tanggapan mengenai kegiatan pendidikan banyak dituangkan dalam media sosial. Untuk mengetahui sentimen tanggapan masyarakat tersebut perlu dilakukan analisis sentimen dengan algoritma machine learning. Metode yang digunakan dalam penelitian ini adalah Support Vector Machine, Naïve Bayes Classifier, dan Lexicon Based. SVM digunakan untuk proses klasifikasi analisis sentimen dan mencari nilai akurasi terbaik menggunakan kernel linear. Naïve Bayes digunakan untuk proses klasifikasi analisis sentimen dan sebagai metode perbandingan terhadap SVM. Lexicon Based digunakan untuk menentukan kelas sentimen positif, netral dan negatif pada data. Hasil penilaian dari 3522 data tweet diperoleh 766 tweet (21.7%) positif, 254 tweet netral (7.2%), dan 2502 tweet negatif (71%). Metode klasifikasi SVM memiliki tingkat akurasi sebesar 83% , presisi sebesar 78%, dan recall sebesar 55. Sedangkan metode Naïve Bayes memiliki tingkat akurasi sebesar 73%, nilai presisi 58% dan recall sebesar 34%. Berdasarkan hasil analisis sentimen ini, dapat disimpulkan bahwa performa metode SVM lebih baik dalam mengklasifikasi data dibandingkan Naïve Bayes. ; After going through the Covid-19 pandemic, various sectors in human life experienced many changes, including in educational activities. This causes various opinions or responses regarding educational activities to be poured on social media. To find out the sentiment of the public's response, it is necessary to carry out sentiment analysis with a machine learning algorithm. The methods used in this research are Support Vector Machine, Naïve Bayes Classifier, and Lexicon Based. SVM is used for the sentiment analysis classification process and to find the best accuracy value using a linear kernel. Naïve Bayes is used for the sentiment analysis classification process and as a comparison method against SVM. Lexicon Based is used to determine the positive, neutral and negative sentiment classes in the data. The results of the assessment of 3522 tweet data obtained 766 positive tweets (21.7%), 254 neutral tweets (7.2%), and 2502 negative tweets (71%). The SVM classification method has an accuracy rate of 83%, a precision of 78%, and a recall of 55. Meanwhile the Naïve Bayes method has an accuracy rate of 73%, a precision value of 58% and a recall of 34%. Based on the results of this sentiment analysis, it can be concluded that the performance of the SVM method is better in classifying data than Naïve Bayes.
| Item Type: | Thesis (S1) |
|---|---|
| Additional Information: | ID Sinta Dosen Pembimbing: Dadan Dasari: 6000619 Lukman: 6675529 |
| Uncontrolled Keywords: | Analisis Sentimen, Komentar, Opini, Twitter, Support Vector Machine, SVM, Naïve Bayes, Lexicon Based, Kuliah, Luring ; Sentiment Analysis, Comments, Opinion, Twitter, Support Vector Machine, SVM, Naïve Bayes, Lexicon Based, Lecture, Offline |
| Subjects: | Q Science > QA Mathematics Q Science > QA Mathematics > QA76 Computer software |
| Divisions: | Fakultas Pendidikan Matematika dan Ilmu Pengetahuan Alam > Jurusan Pendidikan Matematika > Program Studi Matematika (non kependidikan) |
| Depositing User: | Zahra Agusfiyanti Nurlaila |
| Date Deposited: | 03 Sep 2023 20:01 |
| Last Modified: | 03 Sep 2023 20:01 |
| URI: | http://repository.upi.edu/id/eprint/101281 |
Actions (login required)
 |
View Item |