ANALISIS SENTIMEN MAHASISWA TERHADAP HYBRID LEARNING MENGGUNAKAN ALGORITMA NAÏVE BAYES: STUDI KASUS DI UIN RADEN INTAN LAMPUNG

Annisa, Zulfi Rofiatus Sabila (2026) ANALISIS SENTIMEN MAHASISWA TERHADAP HYBRID LEARNING MENGGUNAKAN ALGORITMA NAÏVE BAYES: STUDI KASUS DI UIN RADEN INTAN LAMPUNG. Diploma thesis, UIN RADEN INTAN LAMPUNG.

	PDF Download (7MB)
	PDF Restricted to Repository staff only Download (7MB)

Abstract

ABSTRAK Universitas Islam Negeri Raden Intan Lampung merupakan salah satu perguruan tinggi yang telah mengadopsi metode pembelajaran hybrid semenjak masa Covid-19. Dengan masa implementasi metode pembelajaran yang telah berlangsung cukup lama, maka dilakukannya evaluasi opini mahasiswa terhadap hybrid learning yang dibagi menjadi empat indikator dengan melakukan analisis sentimen menggunakan algoritma naive bayes. Penelitian ini dilakukan kepada 394 mahasiswa dengan pengambilan opini menggunakan kuesioner terbuka yang didapatkan 1.182 opini pada setiap indikator. Proses analisis mencangkup pengambilan data, prapemprosesan data, pelabelan data, ekstraksi fitur, pembagian data, pelatihan model dengan data training, klasifikasi dengan data testing, serta evaluasi model. Hasil pengujian kinerja naive bayes dalam mengklasifikasi sentimen opini mahasiswa memiliki rata-rata persentase akurasi sebesar 80,22%. Hasil sentimen didapatkan bahwa mayoritas mahasiswa bersentimen positif terhadap hybrid learning dengan rata-rata persentase sebesar 70,33%, dengan sentimen positif paling banyak terdapat pada indikator 1 aspek motivasi belajar. Dengan keempat indikator memiliki tingkat sentimen positif lebih tinggi, maka dapat diasumsikan bahwa penerapan hybrid learning di UIN Raden Intan Lampung telah berjalan cukup baik. Namun tetap perlu diadakannya perbaikan lebih lanjut mengingat masih adanya opini negatif, serta penurunan opini positif pada indikator aspek interaksi dan komunikasi, dan aksesibilitas platform pembelajaran. Kata Kunci : Hybrid Learning, Analisis Sentimen, Algortima Naive Bayes ABSTRACT Raden Intan State Islamic University of Lampung is one of the higher education institutions that has adopted a hybrid learning method since the onset of the COVID-19 pandemic. Given that this learning method has been in place for quite some time, an evaluation of student opinions regarding hybrid learning was conducted, divided into four indicators, using sentiment analysis based on the Naive Bayes algorithm. This study involved 394 students, with opinions collected via an open-ended questionnaire, yielding 1,182 responses for each indicator. The analysis process included data collection, data preprocessing, data labeling, feature extraction, data splitting, model training using the training set, classification using the test set, and model evaluation. The results of testing the Naive Bayes model’s performance in classifying student sentiment showed an average accuracy rate of 80.22%. The sentiment analysis revealed that the majority of students held positive sentiments toward hybrid learning, with an average percentage of 70.33%, and the highest concentration of positive sentiment was found in Indicator 1: learning motivation. Since all four indicators showed higher levels of positive sentiment, it can be assumed that the implementation of hybrid learning at UIN Raden Intan Lampung has been quite successful. However, further improvements are still needed, given the continued presence of negative feedback and the decline in positive feedback regarding the indicators for interaction and communication, as well as the accessibility of the learning platform. Keyword : Hybrid Learning, Sentiment Analysis, Naive Bayes Algorithm

Actions (login required)

View Item