MUNANDAR, Adi Ariyo and Farikhin, Farikhin and Widodo, Catur Edi (2023) SENTIMEN ANALISIS APLIKASI BELAJAR ONLINE MENGGUNAKAN SUPPORT VECTOR MACHINE DAN ARTIFICIAL NEURAL NETWORK. Masters thesis, UNIVERSITAS DIPONEGORO.

	Text cover Tesis.pdf Download (1MB)
	Text Bab 1.pdf Download (142kB)
	Text Bab 2.pdf Download (443kB)
	Text Bab 3.pdf Restricted to Repository staff only Download (359kB)
	Text Bab 4.pdf Restricted to Repository staff only Download (506kB)
	Text Bab 5.pdf Download (118kB)
	Text Daftar Pustaka.pdf Download (170kB)
	Text Lampiran.pdf Restricted to Repository staff only Download (244kB)

Abstract

Sentimen analisis merupakan teknik analisa yang digunakan untuk menganalisis berbagai macam data berupa ulasan maupun komentar. Teknik sentimen digunakan untuk menarik sebuah informasi yang didasarkan pada data aplikasi bimbingan belajar online seperti Ruang Guru, Zenius dan Quipper. Aplikasi tersebut tersedia di google play store. Data dikumpulkan dengan menggunakan teknik scraping dan data berupa ulasan yang terdapat di bagian halaman google play store. Data diolah dengan menggunakan algoritma SVM dan ANN. SVM digunakan untuk mengklasifikasikan data ke sentimen netral dan tidak netral. Kemudian data, diolah ke dalam ANN dengan menggunakan algoritma LSTM, untuk menghasilkan klasifikasi positif dan negatif. Selain klasifikasi data juga diolah berdasarkan sentimen aspek, yaitu aspek User interface, User Experience, Functionality dan Performance, dan Support dan Updates. Hasil dari sentimen analisis berbasis aspek yaitu algoritma SVM 86% dan algoritma ANN berbasis LSTM 86%. Aspek user interface menghasilkan precision 0% dan recall 0%. Aspek user experience mendapatkan precision 88% dan recall 74%. Aspek functionality dan performance mendapatkan nilai precision 47% dan recall 60%. Kemudian, aspek support dan updates mendapatkan nilai precision 70% dan recall 62%.
Kata Kunci: sentimen analisis, SVM, LSTM, Aspek Based

Sentiment analysis is an analytical technique used to analyze various kinds of data in the form of reviews and comments. The sentiment technique is used to retrieve information based on data from online tutoring applications such as Ruang Guru, Zenius, and Quipper. Such applications are available on the Google Play Store. Data was collected using scraping techniques and in the form of reviews found on the Google Play Store page. Data is processed using the SVM and ANN algorithms. SVM is used to classify data into neutral and non-neutral sentiments. Then the data is processed into an ANN using the LSTM algorithm to produce positive and negative classifications. In addition to data classification, it is also processed based on sentiment aspects, namely aspects of the User interface, User Experience, Functionality and Performance, and Support and Updates. The results of the aspect-based sentiment analysis are the SVM algorithm (86%), and the LSTM-based ANN algorithm (86%). Aspects of the user interface produce 0% precision and 0% recall. The user experience aspect gets 88% precision and 74% recall. The functionality and performance aspects get a precision value of 47% and a recall of 60%. Then, the aspects of support and updates get a precision value of 70% and a recall of 62%.
Keyword : sentiment analysis, SVM, LSTM, aspect based

Actions (login required)

View Item