HENDRAWAN, ACHMAD ZAKI (2026) ANALISIS SENTIMEN ULASAN MOBILE LEGENDS MENGGUNAKAN TIGA ALGORITMA MACHINE LEANING DAN FEATURE EXTRACTION LATENT DIRICHLET ALLOCATION. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
COVER.pdf Download (492kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
BAB 1.pdf Restricted to Registered users only Download (115kB) |
|
|
Text (BAB II)
BAB 2.pdf Restricted to Registered users only Download (415kB) |
|
|
Text (BAB III)
BAB 3.pdf Restricted to Registered users only Download (889kB) |
|
|
Text (BAB IV)
BAB 4.pdf Restricted to Registered users only Download (1MB) |
|
|
Text (BAB V)
BAB 5.pdf Restricted to Registered users only Download (692kB) |
|
|
Text (DAFTAR PUSTAKA)
DAFTAR PUSTAKA.pdf Restricted to Registered users only Download (703kB) |
|
|
Text (LAMPIRAN)
LAMPIRAN.pdf Restricted to Registered users only Download (962kB) |
Abstract
Sentiment analysis of user reviews for the mobile game application Mobile Legends is essential for understanding user satisfaction levels. This study aims to examine the sentiment of Mobile Legends user reviews available on the Google Play Store in order to assess user satisfaction and identify aspects of the game that require improvement. The research data were collected from user reviews and subsequently processed through text preprocessing stages, including data cleaning, tokenization, and normalization, to ensure data readiness for analysis. The user reviews were then transformed into numerical representations using feature extraction methods such as Term Frequency–Inverse Document Frequency (TF-IDF), Word2Vec, and Bidirectional Encoder Representations from Transformers (BERT). Sentiment classification was performed using several machine learning algorithms, namely Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Naive Bayes. To evaluate the performance of each model, accuracy, precision, recall, and F1-score metrics were employed. The experimental results indicate that the combination of TF-IDF and the SVM algorithm achieved the best performance, with an accuracy rate of 84.7%, and demonstrated strong capability in effectively identifying negative reviews. The best-performing model was then utilized in a subsequent stage, namely topic modeling using Latent Dirichlet Allocation (LDA). The topic modeling results reveal that Mobile Legends user reviews are predominantly focused on three main topics: Graphics and Performance, Matchmaking System, and Game Updates and Changes. The findings of this study are expected to serve as a basis for developers to improve game quality and to contribute as a reference for future research in the field of sentiment analysis for mobile games. Keywords: Analisis Sentimen, Mobile Legends, Support Vector Machine, Word2Vec, Latent Dirichlet Allocation, Ulasan Pengguna Analisis sentimen terhadap ulasan pengguna aplikasi game Mobile Legends sangat penting untuk mengetahui seberapa puas pengguna. Penelitian ini bertujuan untuk mengkaji sentimen ulasan pengguna aplikasi game Mobile Legends yang tersedia di Google Play Store guna mengetahui tingkat kepuasan pengguna serta mengidentifikasi aspek permainan yang masih perlu ditingkatkan. Data penelitian diperoleh dari ulasan pengguna yang selanjutnya melalui tahapan prapemrosesan teks, meliputi pembersihan data, tokenisasi, dan normalisasi, agar data siap untuk dianalisis. Setelah itu, ulasan pengguna dikonversi ke dalam bentuk numerik dengan menerapkan metode ekstraksi fitur Term Frequency–Inverse Document Frequency (TF-IDF), Word2Vec, dan Bidirectional Encoder Representations from Transformers (BERT). Klasifikasi sentimen dilakukan dengan menggunakan beberapa algoritma pembelajaran mesin, yaitu Support Vector Machine (SVM), KNearest Neighbor (KNN), dan Naive Bayes. Untuk mengetahui performa masingmasing metode, dilakukan evaluasi menggunakan metrik akurasi, presisi, recall, dan F1-score. Hasil pengujian menunjukkan bahwa kombinasi metode TF-IDF dan algoritma SVM menghasilkan kinerja terbaik dengan tingkat akurasi sebesar 84,7% serta mampu mengidentifikasi ulasan negatif secara efektif. Model dengan performa terbaik tersebut kemudian digunakan pada tahap lanjutan, yaitu pemodelan topik menggunakan Latent Dirichlet Allocation (LDA). Hasil pemodelan topik menunjukkan bahwa ulasan pengguna Mobile Legends didominasi oleh tiga topik utama, yaitu Grafik dan Performa, Sistem Matchmaking, serta Update dan Perubahan Game. Hasil penelitian ini diharapkan dapat menjadi dasar evaluasi bagi pengembang dalam meningkatkan kualitas permainan, serta memberikan kontribusi bagi penelitian selanjutnya di bidang analisis sentimen pada game mobile. Kata Kunci : Analisis Sentimen, Mobile Legends, Support Vector Machine, Word2Vec, Latent Dirichlet Allocation, Ulasan Pengguna
Actions (login required)
 |
View Item |
