FIRDAUS, ALDITO FAISAL (2022) Analisis Sentimen Pada Ulasan Pelanggan Marketplace Online di Indonesia Menggunakan Algoritma Long Short-Term Memory (LSTM). S1 thesis, Universitas Mercu Buana.
|
Text (HAL COVER)
01 COVER.pdf Download (3MB) | Preview |
|
|
Text (ABSTRAK)
02 ABSTRAK.pdf Download (24kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB 1)
03 BAB 1.pdf Restricted to Registered users only Download (105kB) |
|
|
Text (BAB 2)
04 BAB 2.pdf Restricted to Registered users only Download (142kB) |
|
|
Text (BAB 3)
05 BAB 3.pdf Restricted to Registered users only Download (207kB) |
|
|
Text (BAB 4)
06 BAB 4.pdf Restricted to Registered users only Download (166kB) |
|
|
Text (BAB 5)
07 BAB 5.pdf Restricted to Registered users only Download (110kB) |
|
|
Text (BAB 6)
08 BAB 6.pdf Restricted to Registered users only Download (151kB) |
|
|
Text (DAFTAR PUSTAKA)
09 DAFTAR PUSTAKA.pdf Restricted to Registered users only Download (88kB) |
|
|
Text (LAMPIRAN)
10 LAMPIRAN.pdf Restricted to Registered users only Download (129kB) |
Abstract
The online marketplace is a place for buying and selling activities that are currently popular with the public because of the various conveniences it offers. Along with population growth in Indonesia, the number of online Marketplace customers in Indonesia continues to increase. Many things are considered by customers in using an online marketplace, one of which is customer opinions on product reviews in the marketplace. Therefore, this study aims to provide a user perspective to e-commerce management that product reviews reviewed by customers can be a parameter in determining strategy. The first step is to collect product review data from online marketplace sites. Then pre-processing of the data includes case folding, cleaning, tokenizing, filtering and stemming. Next is the process of labeling the data. The result of the data labeling process is the creation of two data classes, namely the positive data class and the negative data class. Data that already has a data label is used as training data for model classifiers. Then this data is compared with three algorithms, namely, Naive Bayes Classifier, K-Nearest Neigbor (KNN), and Long Short Term Memory (LSTM). The result shows that the LSTM algorithm model has the greatest accuracy value compared to other algorithms, which is 92%. Key words: Sentiment Analysis, Classification, Long Short Term Memory, Online Marketplace online merupakan tempat kegiatan jual beli yang saat ini sedang digemari masyarakat karena berbagai kemudahan yang ditawarkannya. Seiringan dengan pertumbuhan penduduk di Indonesia, jumlah pelanggan Marketplace online di Indonesia terus mengalami kenaikan. Banyak hal yang menjadi pertimbangan pelanggan dalam menggunakan suatu marketplace online, salah satunya adalah opini pelanggan terhadap ulasan produk yang ada pada marketplace. Oleh karena itu, penelitian ini bertujuan untuk memberikan perspektif daalm sudut pandang pengguna kepada manajemen marketplace bahwa ulasan produk yang diulas oleh pelanggan dapat menjadi parameter dalam menentukan strategi. Langkah pertama melakukan koleksi data ulasan produk dari situs marketplace online. Kemudian dilakukan pre-processing pada data meliputi case folding, cleaning, tokenizing, filtering dan stemming. Selanjutnya dilakukan proses pemberian label (labelling) pada data. Hasil dari proses labelling data tercipta dua kelas data yaitu kelas data positif dan kelas data negatif. Data yang sudah memiliki label data, digunakan sebagai data pelatihan (training) untuk model machine learning. Kemudian data ini diproses dengan tiga algoritma, yaitu, Naive Bayes Classifier, K-Nearest Neigbor (KNN), dan Long Short Term Memory (LSTM). Hasilnya didapatkan bahwa model algoritma LSTM memiliki nilai akurasi terbesar dibanding dengan algoritma lainnya, yaitu sebesar 92%. Kata kunci : Analisis Sentimen, Klasifikasi, Long Short Term Memory, Marketplace Online, Ulasan Pelanggan
Actions (login required)
 |
View Item |
