ANALISA SENTIMEN TERHADAP PEMBELAJARAN OFFLINE DI ERA NEW NORMAL MENGGUNAKAN ALGORITMA SVM DAN NAÏVE BAYES

FAUZAN, AZHAR (2021) ANALISA SENTIMEN TERHADAP PEMBELAJARAN OFFLINE DI ERA NEW NORMAL MENGGUNAKAN ALGORITMA SVM DAN NAÏVE BAYES. S1 thesis, Universitas Mercu Buana Jakarta.

Preview	Text (HAL COVER) 1. COVER - AZHAR FAUZAN.pdf Download (1MB) | Preview
	Text (BAB I) 2. BAB 1 - AZHAR FAUZAN.pdf Restricted to Registered users only Download (134kB)
	Text (BAB II) 3. BAB 2 - AZHAR FAUZAN.pdf Restricted to Registered users only Download (142kB)
	Text (BAB III) 4. BAB 3 - AZHAR FAUZAN.pdf Restricted to Registered users only Download (434kB)
	Text (BAB IV) 5. BAB 4 - AZHAR FAUZAN.pdf Restricted to Registered users only Download (327kB)
	Text (BAB V) 6. BAB 5 - AZHAR FAUZAN.pdf Restricted to Registered users only Download (201kB)
	Text (DAFTAR PUSTAKA) 7. DAFTAR PUSTAKA - AZHAR FAUZAN.pdf Restricted to Registered users only Download (80kB)
	Text (LAMPIRAN) 8. LAMPIRAN - AZHAR FAUZAN.pdf Restricted to Registered users only Download (253kB)

Abstract

From the recent events where the spread of COVID-19 in Indonesia is getting wider, the Indonesian government has issued policies to prevent the spread. One of them is the policy applied to the field of education, where the teaching and learning process is carried out online or online. Now, Indonesia has carried out the emergency response period, after which the government began to explore a new normal life or New Normal. Because of this, in a press release, the government allowed the implementation of face-to-face learning in the new school year during the Covid-19 pandemic while still paying attention to health protocols. This causes pros and cons in society. In this study, a comparison of the SVM Algorithm with Naive Bayes is carried out in analyzing sentiment regarding the implementation of offline learning in the new normal period based on community tweet data. The data used in this study were 2,708 data that had passed the preprocessing process, automatic labeling, resampling and TF-IDF. The results obtained using SVM precision, recall and F1-Score are 91%, 91%, 91% and 91.5% while with Naïve Bayes the results are 79%, 83%, 79% and 78.5% with a percentage split of 80% :20%. From these results it can be concluded that the SVM algorithm is better than Naive Bayes in analyzing sentiment regarding offline learning in the new normal era. Key words: New Normal, SVM, Naive Bayes Dari peristiwa yang terjadi belakangan ini dimana semakin luasnya penyebaran COVID-19 di Indonesia, pemerintah Indonesia mengeluarkan kebijakankebijakan guna mencegah penyebaran tersebut. Salah satunya adalah kebijakan yang diterapkan pada bidang pendidikan, dimana proses belajar mengajar dilakukan secara online atau daring. Sekarang, Indonesia telah melaksanakan masa tanggap darurat penanganan yang kemudian pemerintah mulai menjajaki kehidupan normal yang baru atau New Normal. Karena hal tersebut pada siaran pers pemerintah memperbolehkan kembali pelaksanaan pembelajaran tatap muka pada tahun ajaran baru di masa pandemi Covid-19 dengan tetap memperhatikan protokol kesehatan. Hal tersebut menyebabkan pro-kontra pada masyarakat. Pada penelitian ini dilakukan perbandingan Algoritma SVM dengan Naive Bayes dalam menganalisa sentimen mengenai pelaksanaan pembelajaran offline di masa new normal berdasarkan data tweet masyarakat. Data yang digunakan pada penelitian ini sebanyak 2.708 data yang telah melewati proses preprocessing, labeling yang dilakukan secara otomatis, resampling dan TF-IDF. Hasil yang didapat menggunakan SVM precision, recall dan F1-Score sebesar 91%, 91%, 91% dan 91.5% sedangkan dengan Naïve Bayes diperoleh hasil sebesar 79%, 83%, 79% dan 78,5% dengan percentage split sebesar 80%:20%. Dari hasil tersebut dapat disimpulkan bahwa Algorirma SVM lebih baik dibandingkan dengan Naive Bayes dalam menganalisa sentimen mengenai pembelajaran offline di era new normal. Kata kunci: New Normal, SVM, Naive Bayes

Actions (login required)

View Item